Apparel Manufacturing in the USA: 2026 Report

Which figures are current, and which describe earlier years?

The BLS industry page reports an extraction date of September 11, 2026. We reviewed it on September 13. The latest monthly figure, latest quarterly count and latest annual analysis are shown separately because they answer different questions. BLS 2; USFIA study and respondent profile 22.

A search for apparel manufacturing in the USA can represent several different needs: custom garment development, cut-and-sew contracting, integrated knitting, decoration of an existing garment or coordination across suppliers. National industry data describe the businesses classified within a statistical boundary. A buyer's specification describes the work required for a particular product. The two perspectives are complementary, but neither substitutes for the other.

This report addresses three research questions:

1. How did recorded private-sector employment and establishments change between 2021 and 2025?
2. How did that change differ across apparel industry groups and geography, and what does the separate productivity series add?
3. Which supplier-level evidence is still needed before the industry findings can inform an actual sourcing decision?

Official BLS profiles already describe the industry and publish the underlying series. The contribution here is their reproducible comparison, the explicit treatment of classification and disclosure limits, and a framework connecting statistical interpretation to product-specific verification. The study does not claim to discover the industry totals or provide a new directory of factories.

2026 update: recent employment is lower, but this is a different series

BLS's current industry profile reports the following monthly estimates. The measure is apparel employment from Current Employment Statistics, seasonally adjusted, in thousands. July and August are marked preliminary. It provides a timely view of 2026 conditions without changing the longer annual analysis.

Enlarge ↗PNGSVGSource data

The reported May-to-August difference is 0.8 thousand jobs, or a calculated 1.1% decline using May as the denominator. It is a short-window descriptive comparison, not a full-year rate, forecast or causal estimate. Rounding and later revisions may alter the result. Source 2.

Do not calculate an employment change by subtracting this August CES estimate from the 2025 annual QCEW average. Monthly and annual periods, adjustment and statistical programs differ. For sourcing, the current indicator is a reason to verify a particular supplier's staffing and delivery commitments, not evidence that a quoted order cannot be produced.

National employment, 2021 to 2025

We analyzed annual BLS Quarterly Census of Employment and Wages data for 2021 through 2025. The comparison uses private-sector records throughout and separates apparel manufacturing, NAICS 315, from commercial screen printing, NAICS 323113.

Annual average private-sector employment. These columns describe separate industries and should not be added together as a count of apparel jobs. Source: our calculations from the BLS annual industry files 1, with the exact extracts retained in the accompanying data package.

Within this five-year window, both series peaked in 2022. Apparel-manufacturing employment was 12.5% lower in 2025 than in 2021, a reduction of 11,225 jobs. Commercial screen-printing employment finished 2.6% above 2021, but it also fell after 2022 and was below its 2024 level. The endpoint comparison therefore does not describe continuous growth in printing.

Enlarge ↗PNGSVGSource data

Industry-group composition and change

Source extracts: BLS 2021 NAICS 3151 8, BLS 2025 NAICS 3151 9, BLS 2021 NAICS 3152 10, BLS 2025 NAICS 3152 11, BLS 2021 NAICS 3159 12, BLS 2025 NAICS 3159 13.

Annual average private U.S. employment. Percent changes are calculated from unrounded source counts. Group definitions follow NAICS; the 2021 group sum exceeds the parent total by one job. Download group data.

Enlarge ↗PNGSVGSource data

Cut-and-sew remained the largest group, with 60,700 jobs in 2025, equivalent to 77.0% of the parent apparel total. Its employment fell by 10,198 jobs from 2021. Knitting employment fell by 1,867 jobs, while accessories and other apparel gained 839 jobs. The group-level changes sum to a reduction of 11,226, differing by one job from the parent's 11,225 reduction because of the retained baseline reconciliation difference.

These results qualify a general statement that apparel manufacturing contracted: the direction and magnitude differed across groups. The increasing accessories-and-other group does not establish that all its component activities expanded or that it offers the required capability for a particular garment. A group average cannot identify individual winners, closures, technology adoption or reasons for employment change.

Employment, establishments and nominal annual pay

Private U.S. apparel manufacturing, NAICS 315. Annual pay is nominal U.S. dollars. Jobs per establishment is a calculated aggregate ratio. Download both industries and indices.

Enlarge ↗PNGSVGSource data

Across the endpoints, establishments declined by 204, or 3.2%. Employment declined by 12.5%. Average jobs per establishment consequently fell by 9.6%, from 14.07 to 12.72. This arithmetic is consistent with a smaller aggregate employment footprint per recorded establishment, but it does not reveal whether an individual establishment reduced staff. Entry, exit and changes in the establishment mix can also affect the ratio.

Nominal average annual pay rose from $49,701 to $59,722, a 20.2% increase. The data do not establish an equivalent increase in real wages or garment labor costs. Pay and employment describe different aspects of the industry; neither should be substituted for the other when assessing supplier economics.

Geographic change and concentration

California is a large part of the industry, so it is useful to separate its change from the national result. Its annual average apparel-manufacturing employment fell from 28,040 in 2021 to 23,770 in 2025, a reduction of 4,270 jobs, or 15.2%. The rest of the U.S., calculated by subtracting California from the matching national figures, fell from 61,979 to 55,024, a reduction of 6,955 jobs, or 11.2%.

Enlarge ↗PNGSVGSource data

California accounts arithmetically for 38.0% of the national net decline between these endpoints; 62.0% occurred in the rest of the country combined. The comparison shows that the national result is not confined to California. It does not show that every other state declined, or explain why employment changed.

This calculation uses published California and national values. It does not fill in suppressed state observations, and it cannot identify relocation between states. Supplier continuity still requires current, business-specific evidence.

The 2025 state records show a substantial concentration of apparel employment in California. The five largest published state counts are listed below.

Our calculations from BLS 2025 NAICS 315 15. Shares use the published national total, not the sum of visible state records. Figures are rounded independently.

Together, these five published state counts represent 54.2% of national apparel employment. This is a useful description of where activity is recorded. It is not a ranking of quality, price or manufacturing suitability.

Enlarge ↗PNGSVGSource data

Five of the 51 state and District of Columbia employment values are withheld in the 2025 apparel file. We preserved them as unavailable. The difference between the national total and the sum of published state employment is 136 jobs; we did not allocate that difference to particular states or use it to reconstruct protected cells.

Screen printing has a much larger state-level disclosure gap: 24 of 51 employment values are withheld in the 2025 extract. We therefore do not present a nationwide state ranking for that industry. A visible list of states would otherwise look more complete than the underlying data allows.

Productivity, output and hours in 2025

Enlarge ↗PNGSVGSource data

Output per hour increased while output declined because hours declined more sharply. Using the rounded published changes, 0.937 divided by 0.901, minus one, is approximately 4.0%. This is an arithmetic consistency check on the reported measures, not an independent estimate of productivity.

The result should not be interpreted as proof that factories produced more garments or that automation caused the productivity increase. The output measure is not a garment-unit count, and the table alone cannot identify the mechanism behind the change. The simultaneous increase in unit labor costs also shows why rising productivity does not, by itself, establish that labor cost per unit of output fell.

The analysis identifies contraction in employment alongside variation in industrial structure and geography. That variation matters because the capabilities relevant to a buyer are more specific than the parent industry's direction. A search for integrated knitting and a search for cut-and-sew work should not be treated as interchangeable merely because both sit within apparel manufacturing.

The establishment results provide a second distinction. A reduction in employment larger than the reduction in establishments changes the industry-wide ratio of jobs to establishments. It does not show how work is distributed among firms or whether a particular supplier has available time. For an order, the relevant evidence remains a confirmed production schedule, the operations required, staffing or subcontracting arrangements where material to delivery, and an agreed contingency plan.

The state analysis supplies location context without establishing a state-level advantage in quality or cost. California remained the largest published employment location, but most of the national net decline occurred elsewhere. That decomposition does not identify which establishments moved, whether production was relocated, or whether a policy caused an employment change. Those are causal questions requiring additional designs and evidence.

The productivity panel adds information that employment alone cannot provide. Falling employment is not a direct measure of output, and rising output per hour is not necessarily expansion. The 2025 panel demonstrates that productivity and output can move in opposite directions. Supplier assessments should therefore avoid using any one industry indicator as a proxy for product-specific capacity or price.

Finally, production location, decoration location and origin claims remain separate verification tasks. Industry classification cannot establish the origin of a garment sold by a U.S.-based company. A sourcing record should identify who performs each relevant stage, where it occurs, which materials are used and what wording the resulting documentation supports. Appendix A sets out the service categories and links to the official origin guidance; Appendix B translates them into a consistent supplier brief.

Decide which production service you need

A useful sourcing brief starts with the work to be done. A company ordering existing polos with a logo and a brand developing a new polo silhouette may both search for an apparel manufacturer, but they need different capabilities.

These are briefing categories. Supplier terminology and the boundaries of a quoted service can differ, so confirm the actual division of work. Do not assume that “full package” means every component is domestic or every step occurs in the same facility.

BLS classifies apparel manufacturing around cut-and-sew operations and integrated knitting and garment production. Knitting fabric alone belongs to a different subsector. The official classification is useful for interpreting industry statistics, but your product specification still has to identify the exact capability required. BLS apparel-manufacturing overview 2.

For business apparel, ask whether a new garment is necessary to solve the problem. If an existing style meets the fit, fabric and use requirements, decoration may address the need. If the required fit or construction is unavailable, custom development may be justified. This is a purchasing decision to make before comparing prices from businesses offering fundamentally different services.

Manufacturing, decoration and origin are separate questions

A finished garment can pass through several businesses before it reaches the buyer. The address that takes the order, the location that decorates the product and the place where the garment was manufactured may be different.

Commercial screen printing is especially easy to misread in industry statistics. The Census definition includes printing on apparel, but it also includes work on other purchased materials, such as stationery and labels. Consequently, the 74,181 jobs in that category cannot all be described as garment-decoration jobs. Nor does that category capture every possible apparel-decoration activity. Census 2022 NAICS manual, printed page 186 16.

For an order, request a simple process map:

The map is a verification aid. It should describe the proposed product, not a supplier's most domestic product line or a different sample shown on its website.

Read “Made in USA” information carefully

Textile origin labeling has specific rules. FTC guidance distinguishes products made in the U.S. from domestic materials from products made here using imported fabric or yarn; the latter need appropriate qualification. The guidance also explains that imported greige fabric finished in the U.S. does not automatically support an unqualified U.S.-origin label. FTC textile and wool labeling guide 17.

The general FTC Made in USA guide separately addresses textile requirements and mixed domestic/foreign processing. Verify the proposed wording against the actual product and current applicable guidance rather than using a U.S. office address as evidence. FTC Made in USA guidance 18.

Our purchasing recommendation is to obtain the origin documentation and the proposed label or product-page wording before approving production. Ask how substitutions will be handled if fabric, blanks or components change. This section is an orientation to the official guidance, not a determination that a particular product or label complies.

Questions readers ask about U.S. apparel manufacturing

How do I find a clothing manufacturer in the US?

Define the work first: garment development, cut-and-sew production, knitting, finishing or decoration of existing blanks. Use the CFDA production directory, SEAMS search and NIST MEP supplier-scouting route as starting points, then verify the actual facility, product experience, order size and subcontracting arrangements. A directory listing is a lead, not an audit or an endorsement. Sources 19, 20, 21.

How can I find small-batch clothing manufacturers with low minimums?

Send the same specification to potential suppliers and ask where each minimum applies: per style, color, fabric purchase or total order. Ask separately about sample charges, setup, size splits and repeat orders. A low garment minimum can still require a larger fabric commitment. The worked comparison below shows how unit price and total cash requirement can point to different decisions; its example prices are illustrative, not market benchmarks.

How do I contact a manufacturer, and what will production cost?

Provide a brief with quantities, size breakdown, construction, fabric, artwork, origin requirements, delivery point and requested dates. Ask for itemized development, sample, setup, production, freight and inspection costs. There is no defensible universal price for an undefined garment. The filled 200-shirt brief below shows what information to provide and what still needs the supplier's confirmation.

Are clothes printed in the USA also made in the USA?

Printing location alone does not establish garment origin. Confirm where the garment was manufactured and which claim the supplier can substantiate for that specific product. An unqualified U.S.-origin claim has a different standard from a statement about domestic printing or assembly. The FTC's guidance should be checked for the exact labeling and advertising context 17, 18. A brand may offer products with different origins; verify the SKU, not only the company name.

Is U.S. clothing manufacturing growing in 2026?

The monthly CES observations shown here decline from May to August 2026. The separate annual QCEW record shows apparel employment lower in 2025 than in 2021. Neither finding is a forecast for the rest of 2026, and neither measures the whole U.S. clothing retail market. Our source selection does not support a claim that every apparel category or manufacturer is shrinking. BLS 1, 2.

Do 2026 sourcing surveys prove that production is returning to the USA?

No single executive survey establishes a national change in domestic production. USFIA's 2026 study surveyed 30 companies in April to June; its respondent profile is concentrated in large businesses. It offers current evidence about sourcing decisions and business expectations. Our QCEW analysis describes recorded domestic employment, while CES provides a monthly indicator. These sources are complementary, but their populations and outcomes differ. A causal claim about reshoring would need a separate design that tracks actual production, location and timing. USFIA 22.

Where to begin a U.S. supplier search

Start with a directory or referral route that lets you look for the actual work required. Treat the result as a candidate list, then verify the businesses individually.

The CFDA Production Directory is an open-access resource for finding U.S. fashion manufacturing services. It offers filters that include service, product and production minimums. Use the filters to narrow the brief, then check the company's current capabilities and terms directly. Directory inclusion is not an endorsement from this report. CFDA Production Directory 19.

SEAMS, the U.S. sewn-products industry association, offers its own search resource. It can be another starting point for identifying businesses in the wider production network. Check whether a result is a garment contractor, material supplier, equipment business or another kind of service before adding it to your shortlist. SEAMS Power Search 20.

For an industrial sourcing problem, the NIST MEP supplier-scouting program describes a process for connecting domestic manufacturing needs with U.S. suppliers. Ask about applicability to your situation rather than assuming that every apparel order qualifies or that an introduction guarantees a match. NIST supplier scouting 21.

Build a shortlist around a consistent brief. Record who the company is, which stage it performs, the stated production location and the evidence still needed. A website that says “U.S.-based” has answered a different question from one that identifies the facility proposed for the work.

Avoid ranking candidates solely by the number of services on their websites. A narrow specialist may be appropriate when you already have a production-ready specification. A coordinating partner may be more useful when several stages still need to be developed. The best fit depends on the work, responsibilities and evidence required for the order.

Prepare a brief suppliers can actually quote

The following is our recommended request-for-quote structure. It is intended to reduce avoidable differences between proposals, not to impose one industry's universal document format.

Describe the product and its use. Identify the garment, intended wearer, work environment and must-have features. State whether you are matching an existing approved item or developing something new. Separate requirements from preferences so the supplier can flag the points that drive cost or feasibility.

State the quantity in a usable form. Include the total and the split by style, color and size. Explain the expected first order and the assumptions behind a reorder. Do not present an aspirational annual forecast as a committed purchase unless you intend it to be one.

Supply the technical information you already have. This may include measurements, construction details, materials, artwork and an approved reference sample. List what is missing. Ask the supplier to quote development work explicitly rather than assume the production price includes it.

Assign material responsibilities. Identify who sources fabric, trims, labels and packaging, who approves alternatives, and who owns unused material. If the buyer supplies inputs, agree how quantity, defects and delivery timing will be checked before production begins.

Describe quality acceptance. Define the approved reference, key measurements, tolerances and intended care process. Identify which checks are required at sample, pre-production and finished-order stages. Where formal performance requirements apply, put the relevant specification in the brief rather than substituting an informal description such as “heavy duty.”

State the dates and what they mean. A needed delivery date is different from a factory completion date. Ask for separate milestones for material availability, samples, approval, production, inspection and dispatch. Make clear which milestones depend on your response.

Request traceability and origin evidence. Specify the information needed about garment construction, decoration, subcontracting and any intended origin claim. Ask the supplier to notify you before changing a stage or source that affects the approved specification.

A clear brief does not remove every uncertainty. It makes uncertainty visible early enough to price, resolve or reject it. Keep all candidates on the same version of the brief, and distribute material changes consistently when comparing quotations.

A filled brief: 200 work shirts, with decisions still open

This hypothetical example shows how a buyer can make missing information visible. It is not a supplier quotation or an Arklavo order. The quantities, material preference and requested production route are invented for instruction. No delivery date is assumed, and the desired U.S. construction requirement is not an origin claim.

The example exposes an immediate choice: the buyer needs development work before a production quote can become firm. Asking only for a price on 200 shirts would conceal the missing pattern, size split, fabric specification and acceptance criteria. Ask candidates to identify which development stages they can perform, what each stage costs and which decisions the buyer must make first.

Treat the U.S. construction requirement as a question to verify for this product. Embroidery location alone does not answer it. The buyer should also request the origin records and proposed label wording discussed earlier, then resolve those details before approving a customer-facing claim.

The filled brief CSV contains this example. Replace its assumptions with your actual requirements before using it alongside the blank supplier template. A useful first response identifies the unresolved inputs and assigns the next action; it need not pretend every term is already settled.

Minimum orders, lead times and prices need a defined denominator

There is no single minimum order, lead time or garment price that this research can establish for U.S. apparel manufacturing. The BLS data does not measure those commercial terms, and a quote for one supplier's product is not an industry average.

Ask what a minimum applies to: the whole order, each style, each color, each fabric or another unit. A headline minimum can look compatible with a project until the size and color split is discussed. If your order falls below the relevant threshold, ask whether the supplier can offer a different input, a smaller development run or another defined scope.

For timing, ask what starts the clock. Does the schedule begin when the deposit arrives, when fabric is received or when the final sample is approved? Ask whether the proposed production slot is reserved and what happens if an approval or input is late. A shorter geographic distance does not resolve unfinished development work.

For price, compare the complete scope. Record garment or production cost, development, sampling, labels, decoration, inspection, packaging and delivery separately where possible. Identify excluded costs and the circumstances that trigger a revised quote. Confirm the treatment of defects, rework and shortages before the order is placed.

A worked example: cheaper per unit can require more cash

Consider a constructed example for a buyer who currently needs 200 garments. These numbers are illustrative arithmetic, not market prices or supplier quotations.

Option B has a lower all-in cost per garment ordered, but requires $4,700 more initially and leaves 300 garments beyond the current requirement. Those extra units may have substantial future value. They are not automatically waste, and the table does not assume they will be unused.

The decision turns on the likelihood and timing of future need, whether the product will remain suitable, and the value of keeping cash available. A lower unit price answers only part of the question. Use your own demand and inventory assumptions before deciding whether the larger order is economical.

The same approach applies to domestic and overseas proposals. Compare like-for-like scope, delivery, development and risk assumptions. Do not use an unsupported industry-wide cost premium or promise a universal domestic lead-time advantage.

Verify the sample, then the production process

An attractive sample is useful evidence about that sample. It does not, by itself, establish repeatable bulk production, the source of every future input or the supplier's capacity on your required dates.

Use a sequence of approvals that fits the project. A development sample can resolve the concept and fit. A pre-production sample can establish the agreed materials and construction. Finished-order checks can establish whether delivered goods match that reference. If the project combines garment manufacture and decoration, ensure the approved reference covers both.

For business apparel, include actual use in the evaluation. Check movement, fit across the intended size range, visibility and placement of branding, and the care process the organization expects to use. Record what was checked and the results. Do not label an informal wear check a certified performance test.

Ask who will perform the work shown in the sample. If an intermediary or subcontractor is involved, clarify the communication and approval path. Subcontracting is not inherently evidence of poor quality; an undisclosed change in who performs an agreed stage is a different issue. Focus on the documented process and the supplier's ability to meet the specification.

Before approving the order, settle the acceptance standard and response to problems. Decide who investigates a discrepancy, how rework or replacements are authorized, and which records establish what was approved. A clear record is more useful than relying on recollections after a delivery problem arises.

Plan the reorder while placing the first order

A first order can satisfy a launch and still leave a poor replenishment process. Ask which inputs and production details will be retained for a repeat order, what can change, and whether the reorder has a different minimum or schedule.

Keep the approved product specification, size breakdown, artwork version, sample reference and relevant origin documentation together. Record substitutions and the date they were approved. If the business changes its logo, garment range or use requirements, evaluate the effect on stock before repeating the previous order automatically.

Consider continuity at the stage where it matters. Another decorator may be able to apply an approved design to an available garment, while another custom manufacturer may need its own sampling and setup. Do not assume a backup supplier is interchangeable until the relevant work has been checked.

The employment data in this report provides industry context for these decisions. It cannot forecast which businesses will remain available or identify a particular supply interruption. The actionable step is to understand the order's dependencies and keep the records needed to move deliberately if conditions change.

Use one brief to compare suppliers

The downloadable apparel supplier brief turns the purchasing framework into fields you can reuse. Complete the product requirement and evidence requested before circulating it. Give each supplier the same quantity, size breakdown, material specification, decoration requirement and required delivery date. Record any proposed change as a deviation rather than quietly treating it as an equivalent quote.

Separate a supplier's answer from the evidence supporting it. An asserted production location, an approved sample and a dated quotation answer different questions. Keep the source document and the date with each response so a later reviewer can see which decision it supports.

The worked quote comparison is also available as a CSV. Its prices are invented to demonstrate the arithmetic; they are not market benchmarks. Replace them with actual, comparable proposals before using the sheet for a purchasing decision.

The study uses administrative secondary data available on September 13, 2026. Series may be revised, and a five-year window cannot establish the full historical trajectory of domestic apparel production. No current monthly employment series has been joined to the annual QCEW observations.

QCEW records establishments and employment within its coverage. It is not a complete count of self-employed makers, independently owned factories, garments produced or unique people working during a year. Establishment and industry aggregates cannot resolve individual supplier availability, specialization or commercial reliability.

Classification changes were checked at the boundaries needed for this analysis. That check does not remove administrative reclassification or revision risk. The one-job 2021 parent/group difference remains visible. The study does not reconstruct suppressed state cells or treat unavailable values as zero.

The comparison with commercial screen printing is contextual. That industry covers non-apparel products as well, and it does not encompass every apparel-decoration activity. Adding the two series would create a misleading total of apparel employment.

Nominal pay does not account for inflation or workforce composition. The separate productivity panel concerns annual rates and uses its own measurement framework. No employment-to-output conversion, factory-capacity estimate, tariff effect or causal policy conclusion is calculated.

The practical examples are authored tools, including an explicitly hypothetical quote comparison. No supplier survey, factory inspection, garment test or quote-collection study was conducted. A future supplier-level study could compare verified capabilities and quote components against a standardized brief, but it would require new observations and consent where appropriate. This report does not present such evidence as already collected.

This 2026 report combines a current-conditions review through September 13 with a reproducible annual analysis. The latest monthly observation reviewed is August 2026 and is preliminary; the annual comparisons remain explicitly 2021 to 2025.

U.S. apparel manufacturing recorded fewer private-sector jobs and establishments in 2025 than in 2021, with substantial differences across industry groups and a net employment decline extending beyond California. The separate productivity series shows why output, hours and output per hour require their own interpretation.

For research, the useful result is a reproducible description with explicit boundaries and unresolved cells. For sourcing, the useful next step is a product-specific brief and verified evidence about capability, schedule and origin. Industry context can improve the questions asked of a supplier; it cannot answer those questions on the supplier's behalf.

Data selection and definitions

For the quantitative analysis, we downloaded the annual QCEW industry slices for NAICS 315 and 323113 for 2021 through 2025 on September 13, 2026. We selected private ownership code 5, total establishment-size code 0 and annual records. National results use only the US000 national row. State results use the matching state industry aggregation level for the 50 states and District of Columbia, excluding territories, counties, metropolitan areas and other geographic records. BLS data-slice documentation 1, aggregation codes 3.

The 2021 data uses the earlier industry classification, and BLS introduced NAICS 2022 with first-quarter 2022 data. We checked the official concordance in both directions. All seven mappings touching apparel manufacturing remained within the 315 subsector, and commercial screen printing mapped unchanged from 323113 to 323113. This supports the industry-boundary comparison used here; it does not remove the possibility of revisions or establishment reclassification in administrative data. BLS transition notice 4, Census 2022-to-2017 concordance 5.

Employment change is calculated as the difference between the endpoint annual averages divided by the baseline. State shares divide published state employment by the matching national total. We did not mix the two industries, sum multiple geographic levels or convert suppressed values into zeros. Five state apparel values and 24 state screen-printing values were withheld in the 2025 extracts. State comparisons explicitly identify published values.

QCEW is an establishment-based employment and wage source with its own coverage. It is not a count of all self-employed makers or a directory of independently owned factories. Annual average jobs are not production volume, the number of distinct people employed during the year or spare capacity. The data files reflect the version available on the observation date and may be revised; we do not label every 2025 observation final. Newer monthly employment estimates from other BLS programs are a different series and were not spliced into these annual results. QCEW overview 6, BLS revision policy 7.

Industry-group extension

We additionally extracted annual private U.S. records for industry groups 3151, 3152 and 3159 in 2021 and 2025. The national group filter uses area US000, ownership code 5, size code 0 and aggregation level 16. The seven relevant rows in the official 2022-to-2017 concordance preserve their four-digit group boundaries, supporting this comparison at the group level. It does not imply that every six-digit code remained unchanged. Census concordance 5.

The three 2021 group employment counts sum to 90,020, one job more than the published 90,019 parent total. The 2025 group sum matches the parent total of 78,794. We retain that one-job reconciliation difference and do not assign a cause or adjust any source observation to eliminate it. Group shares use the corresponding published parent total; endpoint changes use each group's own baseline.

Establishments, pay and indices

The employment and establishment indices set each series' 2021 value to 100. Average jobs per establishment is total annual average employment divided by annual average establishments. This ratio describes an aggregate mean, not the median establishment, number of businesses or production-line capacity.

Annual pay is reported in nominal U.S. dollars. The change calculation does not adjust for inflation or changes in the mix of occupations, establishments or hours. It cannot establish a change in real purchasing power, a sewing-operator wage or the labor component of a specific supplier quote.

Separate productivity measures

We transcribed the 2025 annual changes published in BLS's apparel Industry at a Glance table, whose data extraction date is September 11, 2026. Its productivity, output, hours and unit-labor-cost measures are a separate statistical series. They are not derived from the QCEW employment counts in this study, and they have not been spliced into the employment trend. The source table and figure data are retained for checking. BLS industry profile 2.

This 2026 report reviews source information available through September 13, 2026. U.S. apparel-manufacturing statistics provide a basis for understanding industrial activity, but they do not directly establish supplier capability, garment origin or available capacity. This report analyzes annual private-sector employment, establishments and nominal annual pay in BLS Quarterly Census of Employment and Wages data for 2021 through 2025. It compares apparel manufacturing with commercial screen printing, examines apparel industry groups and published state counts, and considers separate BLS productivity measures for 2025. Apparel employment declined from 90,019 to 78,794 jobs between the endpoints, a reduction of 12.5%, while establishments declined by 3.2%. The contraction was heterogeneous: apparel knitting and cut-and-sew employment declined, while accessories and other apparel employment increased by 7.3%. California accounted for 38.0% of the national net employment decline; the remainder occurred outside California. Separate productivity data show that output per hour rose by 4.0% in 2025 while output fell by 6.3%. These results support a distinction between employment, output, productivity and the capabilities required for a particular sourcing brief. A separate current-conditions panel records BLS seasonally adjusted employment of 72.0 thousand in August 2026, marked preliminary. The monthly estimate is not joined to the annual QCEW series. The report provides reproducible extracts and practical verification templates; it does not identify causal drivers or evaluate individual manufacturers.

September 13, 2026. The Citation, reuse and data files (ZIP) provides source-linked data, editable SVGs, PNGs, analysis code, templates and a suggested citation. The claim ledger separates source observations, calculated results and interpretations.

Arklavo commissioned this research. The report uses primary-source checks, reproducible calculations and separate internal editorial review. No external academic peer review is implied. The analysis establishes neither that Arklavo owns a U.S. factory nor that a particular Arklavo product is U.S.-made; those claims require product-specific evidence.

Dates marked n.d. are not asserted publication dates. Observation periods are specified in the text and data files.

1. U.S. Bureau of Labor Statistics. (n.d.). QCEW CSV Data Slices. Accessed September 13, 2026.

2. U.S. Bureau of Labor Statistics. (n.d.). Industries at a Glance: Apparel Manufacturing, NAICS 315. Accessed September 13, 2026.

3. U.S. Bureau of Labor Statistics. (n.d.). QCEW Aggregation Level Titles. Accessed September 13, 2026.

4. U.S. Bureau of Labor Statistics. (n.d.). NAICS 2022 transition. Accessed September 13, 2026.

5. U.S. Census Bureau. (2022). 2022 NAICS to 2017 NAICS Concordance. Accessed September 13, 2026.

6. U.S. Bureau of Labor Statistics. (n.d.). QCEW Overview. Accessed September 13, 2026.

7. U.S. Bureau of Labor Statistics. (n.d.). QCEW Revisions. Accessed September 13, 2026.

8. U.S. Bureau of Labor Statistics. (n.d.). QCEW annual industry slice, 2021, NAICS 3151. Accessed September 13, 2026.

9. U.S. Bureau of Labor Statistics. (n.d.). QCEW annual industry slice, 2025, NAICS 3151. Accessed September 13, 2026.

10. U.S. Bureau of Labor Statistics. (n.d.). QCEW annual industry slice, 2021, NAICS 3152. Accessed September 13, 2026.

11. U.S. Bureau of Labor Statistics. (n.d.). QCEW annual industry slice, 2025, NAICS 3152. Accessed September 13, 2026.

12. U.S. Bureau of Labor Statistics. (n.d.). QCEW annual industry slice, 2021, NAICS 3159. Accessed September 13, 2026.

13. U.S. Bureau of Labor Statistics. (n.d.). QCEW annual industry slice, 2025, NAICS 3159. Accessed September 13, 2026.

14. U.S. Bureau of Labor Statistics. (n.d.). QCEW annual industry slice, 2021, NAICS 315. Accessed September 13, 2026.

15. U.S. Bureau of Labor Statistics. (n.d.). QCEW annual industry slice, 2025, NAICS 315. Accessed September 13, 2026.

16. U.S. Census Bureau. (2022). 2022 NAICS Manual, commercial screen printing, printed page 186. Accessed September 13, 2026.

17. Federal Trade Commission. (n.d.). Threading Your Way Through the Labeling Requirements Under the Textile and Wool Acts. Accessed September 13, 2026.

18. Federal Trade Commission. (n.d.). Complying with the Made in USA Standard. Accessed September 13, 2026.

19. Council of Fashion Designers of America. (n.d.). Production Directory. Accessed September 13, 2026.

20. SEAMS. (n.d.). Power Search. Accessed September 13, 2026.

21. National Institute of Standards and Technology. (n.d.). Supplier Scouting. Accessed September 13, 2026.

22. United States Fashion Industry Association, Sheng Lu and Emilie Delaye. (2026). 2026 Fashion Industry Benchmarking Study. Respondent profile and methods; accessed September 13, 2026.