U.S. Apparel Supply Chains: The 2026 Import Concentration Report

Executive summary: the latest 2026 findings

U.S. apparel imports in OTEXA's total-apparel category reached $41.835 billion in January-July 2026, compared with $45.610 billion in the same seven months of 2025. The decrease was 8.28%, calculated from the official data exported on September 14, 2026. Vietnam supplied 22.39% of the 2026 value, followed by Bangladesh at 11.14% and China at 10.89%. Together, those three origins accounted for 44.41%. These are nominal import values, not retail sales or garment counts. [1]

The original analysis in this report adds a category-level concentration comparison. Across twelve predefined shirt and trouser categories, the leading origin's share ranged from 13.65% to 39.24%. A national total therefore gives a limited view of the sourcing pattern for an individual uniform program. The same headline apparel market contains much more dispersed cotton knit-shirt trade and more concentrated man-made-fiber trouser trade. [1; calculations in the downloadable concentration table]

Three findings deserve particular attention:

1. A larger lead-country share can coexist with lower overall concentration. Vietnam's share increased by 1.71 percentage points between the matched seven-month periods. Yet the top-three share fell from 46.73% to 44.41%, and the all-origin concentration index also declined.
2. An origin can gain share while its import value falls. Vietnam's value decreased 0.68%, less than the overall market. Its larger share does not establish that U.S. buyers purchased more Vietnamese apparel in dollars, pieces or orders.
3. A long list of origin countries can obscure the distribution of spending. There were 190 origins with positive total-apparel values in the 2026 period. The distribution had the same concentration index as approximately 10.73 equally sized origins. That mathematical comparison is not a count of factories, qualified vendors or practical alternatives.

For buyers, the practical implication is to assess country exposure at the garment category, supplier and order level. Start with a trade benchmark, then establish the actual cutting-and-sewing origin, upstream dependencies, inventory position and approved substitution options for the products being bought. An origin share alone cannot establish delivery reliability or explain why a shipment changed.

Scope in one sentence: this is a reproducible secondary analysis of U.S. general-import values by reported origin, covering OTEXA category 1 and twelve specified commodity categories; it does not estimate imports as a share of U.S. apparel consumption. [1-3]

1. What this report measures

“Where are U.S. clothes made?” can refer to several different questions. It can mean the origin of goods entering the country, the source of products sold by a particular retailer, the location of a brand's factories, or the share of domestic demand supplied by imports. Those questions require different denominators. This report answers the first and develops a method for investigating the third at an individual buyer's scale.

The observation unit is a reported origin, an OTEXA category and a defined time period. The source is the Office of Textiles and Apparel's public Annual Data explorer, which presents U.S. Census Bureau trade statistics. The selected table contains all returned reporting areas, including small and zero-valued rows. It is not the threshold-based Major Shippers report. [1,2]

“Apparel” also needs a boundary. Category 1 is OTEXA's total-apparel aggregate under its textile-category system. Its headnotes identify exclusions, including certain silk apparel categories. It should not be described as an exact total of every item a reader might call clothing. The twelve detailed categories cover cotton and man-made-fiber knit shirts, non-knit shirts, and trousers/breeches/shorts, each split by OTEXA's men/boys and women/girls classifications. They include retail and children's clothing as well as garments that could be used for uniforms. [2,4]

Reading the dates correctly

This is a 2026 report using the latest period displayed by the source on September 14, 2026: July 2026. Its principal comparison is January-July 2026 against January-July 2025. Calendar-year 2024 and 2025 data provide separate historical context. There is no full-year 2026 total in this analysis, and the seven-month values have not been annualized.

The edition date tells a reader when the analysis was assembled. The data-period label tells a reader what happened during the measured interval. The extraction date identifies the source version. Keeping all three visible lets another researcher explain a later discrepancy without assuming that either calculation was fabricated or that every source has been updated at the same time.

2. The 2026 apparel-import snapshot

The ten largest reported origins by value are shown below. Rankings use January-July 2026 values; the same origins' January-July 2025 values remain in the comparison even where their rank changed. The denominator for each share is the World value for category 1 in the same period. [1]

Enlarge figure ↗SVGPNGCSV

Read the data behind this figure (10 rows)

The ten largest apparel import origins in 2026: equivalent figure data

category_id	country	rank_2026	dollars_2025	dollars_2026	change_dollars	change_pct	share_2025_pct	share_2026_pct	share_change_pp	contribution_to_world_growth_pp	scope	source	observed_date
1	Vietnam	1	9429539296	9365565925	-63973371	-0.6784358067963875	20.674134169774582	22.387029550432175	1.7128953806575922	-0.1402607289528778	OTEXA category 1	OTEXA Annual Data	2026-09-14
1	Bangladesh	2	4969379221	4658856821	-310522400	-6.2487161110138185	10.895295043632203	11.136322797595332	0.24102775396312914	-0.6808160567339355	OTEXA category 1	OTEXA Annual Data	2026-09-14
1	China	3	6913609128	4555672975	-2357936153	-34.105719738340404	15.157992158777281	10.889676750999994	-4.268315407777287	-5.169742323632194	OTEXA category 1	OTEXA Annual Data	2026-09-14
1	Indonesia	4	2660961334	2743329320	82367986	3.095422129873008	5.834120831654505	6.557531605161063	0.7234107735065578	0.1805906673065647	OTEXA category 1	OTEXA Annual Data	2026-09-14
1	Cambodia	5	2363611510	2621126266	257514756	10.894969622144037	5.182185465168191	6.265422895131236	1.0832374299630452	0.5645975321932382	OTEXA category 1	OTEXA Annual Data	2026-09-14
1	India	6	3299645190	2453382879	-846262311	-25.64706998087876	7.234426330844079	5.864456611648629	-1.36996971919545	-1.8554183837867007	OTEXA category 1	OTEXA Annual Data	2026-09-14
1	Mexico	7	1534912085	1386247192	-148664893	-9.685564043233134	3.3652734654342598	3.3136232343063265	-0.05165023112793321	-0.32594571672456624	OTEXA category 1	OTEXA Annual Data	2026-09-14
1	Pakistan	8	1341865194	1268895467	-72969727	-5.437932761522988	2.942020833432942	3.0331109239550234	0.09109009052208128	-0.1599851147520816	OTEXA category 1	OTEXA Annual Data	2026-09-14
1	Italy	9	1159093974	1214610410	55516436	4.789640637023966	2.5412974676311495	2.903350353698198	0.36205288606704844	0.12171901621732249	OTEXA category 1	OTEXA Annual Data	2026-09-14
1	Jordan	10	1167201910	1104083505	-63118405	-5.407668070042826	2.559074004897933	2.6391517875712065	0.08007778267327348	-0.1383862278516317	OTEXA category 1	OTEXA Annual Data	2026-09-14

The five largest origins supplied 57.24% of the 2026 value. Vietnam was the largest, but the next four were materially smaller. Bangladesh and China each supplied approximately one-ninth of the aggregate. Indonesia and Cambodia each supplied a little over 6%. These relationships are more informative than a claim that one country “dominates” every apparel category.

The ranking is descriptive. It does not grade quality, labor conditions, environmental performance, certification, lead time or ease of doing business. A large trade share may identify an established flow of goods; it does not show whether a specific factory has the machinery, materials, capacity or documentation a buyer needs.

Import value is not the amount a team pays for uniforms

Customs import value is measured at a different point in the commercial chain from a decorated garment's delivered price. Census guidance describes customs value as generally excluding U.S. import duties, international freight and insurance, among other bringing-to-market charges. Subsequent distribution, decoration and retail activity also sit outside the imported goods value measured here. [3]

For that reason, a $1 increase in customs import value is not a $1 increase in the final price of a uniform. A decline in value may reflect changes in quantities, prices, product mix or the timing of shipments. This report does not divide total dollars by an assumed garment count or use the resulting figure as a wholesale price.

For a buyer requesting a quotation, compare like-for-like specifications and ask which costs are included. A sourcing benchmark can provide context, but an actual quoted product, volume, delivery address and decoration specification determine the purchasing comparison. Keep those commercial inputs separate from national trade values.

3. Why the leading country's share rose while concentration fell

The simplest concentration measure is the share supplied by the largest origin, called CR1 here. CR3 and CR5 sum the three and five largest shares. Those measures are easy to read, but they omit the distribution below their cutoffs. The Herfindahl-Hirschman Index, or HHI, uses every origin share. In this report it is calculated as the sum of squared percentage shares. [Original calculation]

Enlarge figure ↗SVGPNGCSV

Read the data behind this figure (10 rows)

A larger leader, lower broader concentration: equivalent figure data

measure	period	value	unit	scope	observed_date
cr1_pct	January-July 2025	20.674134169774582	percent	OTEXA category 1	2026-09-14
cr3_pct	January-July 2025	46.72742137218407	percent	OTEXA category 1	2026-09-14
cr5_pct	January-July 2025	59.79596853468265	percent	OTEXA category 1	2026-09-14
hhi	January-July 2025	955.0646388129053	index or effective count	OTEXA category 1	2026-09-14
effective_origins	January-July 2025	10.470495496963922	index or effective count	OTEXA category 1	2026-09-14
cr1_pct	January-July 2026	22.387029550432175	percent	OTEXA category 1	2026-09-14
cr3_pct	January-July 2026	44.4130290990275	percent	OTEXA category 1	2026-09-14
cr5_pct	January-July 2026	57.2359835993198	percent	OTEXA category 1	2026-09-14
hhi	January-July 2026	931.6395823801241	index or effective count	OTEXA category 1	2026-09-14
effective_origins	January-July 2026	10.733764632941323	index or effective count	OTEXA category 1	2026-09-14

Between January-July 2025 and January-July 2026, Vietnam's lead-country share increased from 20.67% to 22.39%. At the same time, the top-three share decreased by 2.31 percentage points, the top-five share decreased by 2.56 points, and HHI decreased from 955.06 to 931.64. These measures are not contradictory. They describe different parts of the distribution.

China's share fell substantially, while the shares of several smaller origins rose. Squaring a share gives greater weight to larger positions, so movements among those positions matter more to HHI than the appearance of a tiny new flow. A count of positive origins increased from 177 to 190, but that count alone cannot explain how concentrated the spending was.

What an effective number of origins means

The inverse form, 10,000 divided by HHI, expresses the same distribution as a number of equally sized origins. If four origins each supplied 25%, HHI would be 2,500 and the effective number would be four. If one supplied everything, HHI would be 10,000 and the effective number would be one. This is a mathematical description, not a claim that the real origins are equivalent or substitutable.

The 2026 apparel total had an effective number of 10.73, compared with 10.47 for the same seven months of 2025. Neither number predicts disruption. Two factories in different countries could depend on the same fabric mill; two suppliers in one country could operate independently. These relationships are outside a country-level trade table.

We do not apply merger-control thresholds or label the result “safe” or “unsafe.” Those thresholds concern different entities, markets and policy questions. For a procurement decision, the relevant issue is whether the buyer understands its exposure and can meet the operational requirement if a particular supply route becomes unavailable.

4. Garment category changes the sourcing picture

National apparel totals combine very different products. To test how much the aggregate conceals, this study fixed a twelve-category frame before the main calculation: three garment groups, two fiber groups, and the two official sex/age groupings. No category was selected because it produced an unusually high concentration score.

The detailed categories together provide useful comparisons, but they are not a complete uniform-market classification. A knit shirt category includes products with different weights, constructions and uses. A trouser category includes shorts. The official men/boys and women/girls labels include children, so those labels should remain visible wherever the results are reused. [4]

Enlarge figure ↗SVGPNGCSV

Read the data behind this figure (12 rows)

Category detail changes the sourcing picture: equivalent figure data

category_id	description	period	world_dollars	listed_origins	positive_origins	leader	cr1_pct	cr3_pct	cr5_pct	hhi	effective_origins	observed_date
338	338 - Doz M&B KNIT SHIRTS, COTTON	jan_jul_2026	4027423725	210	123	Vietnam	13.649234958509362	35.645059075575666	49.4708650999964	713.6322270324562	14.01282007902538	2026-09-14
339	339 - Doz W&G KNIT SHIRTS/BLOUSES, COTTON	jan_jul_2026	2919555946	210	131	Vietnam	26.073293647375785	45.40459599057123	60.19919170954637	1123.6777761326912	8.899348382964847	2026-09-14
340	340 - Doz M&B COTTON SHIRTS, NOT KNIT	jan_jul_2026	876271831	205	108	Bangladesh	26.73724142571462	57.72653497519538	68.8430547073012	1343.784417311027	7.441669862499559	2026-09-14
341	341 - Doz W&G COT. SHIRTS/BLOUSES,N-KNIT	jan_jul_2026	560227571	201	103	India	35.94859114850669	57.22277438573262	76.2008664154089	1753.2617973788747	5.703654762198089	2026-09-14
347	347 - Doz M&B COT. TROUSERS/BREECHES/SHORTS	jan_jul_2026	2800401432	201	117	Bangladesh	32.623030739830014	57.57435386142168	71.23548003527803	1547.6342013565684	6.461475193062138	2026-09-14
348	348 - Doz W&G COTTON TROUSERS/SLACKS/SHORTS	jan_jul_2026	3090867922	209	118	Vietnam	21.902618943417927	52.16787296290042	67.90654819187063	1233.3603272581286	8.107930650105232	2026-09-14
638	638 - Doz M&B MMF KNIT SHIRTS	jan_jul_2026	2703731384	181	104	Vietnam	18.05177925175129	37.3560526011189	51.95881348692441	768.7554232390007	13.00803831453566	2026-09-14
639	639 - Doz W&G MMF KNIT SHIRTS & BLOUSES	jan_jul_2026	1756333444	193	113	Vietnam	26.566276101726388	46.22144614812676	60.59681626150256	1113.2360677368888	8.982820706060236	2026-09-14
640	640 - Doz M&B NOT-KNIT MMF SHIRTS	jan_jul_2026	487995994	178	94	Vietnam	21.16848893640713	48.54899874444461	63.53137378418725	1071.762993305897	9.330421056202537	2026-09-14
641	641 - Doz W&G NOT-KNIT MMF SHIRTS & BLOUSES	jan_jul_2026	437353040	191	105	Vietnam	27.94699746456547	56.81086062646323	72.60199243156055	1380.9616437948887	7.2413307385713885	2026-09-14
647	647 - Doz M&B MMF TROUSERS/BREECHES/SHORTS	jan_jul_2026	1908743290	186	107	Vietnam	26.137940372274997	50.18050531038147	63.7505338394667	1170.5643359475514	8.542887983944235	2026-09-14
648	648 - Doz W&G MMF SLACKS/BREECHES/SHORTS	jan_jul_2026	1976486449	197	109	Vietnam	39.236912117073665	59.16728149548776	73.88996457521374	1895.5814017473808	5.2754263102506815	2026-09-14

The leading-origin share ranged from 13.65% in men's/boys' cotton knit shirts to 39.24% in women's/girls' man-made-fiber trousers, slacks and breeches/shorts. In the former category, the five largest origins supplied 49.47% of import value; in the latter, they supplied 73.89%. Those are different distributions despite belonging to the same broad apparel market.

Cotton and man-made-fiber categories

For men's/boys' knit shirts, the effective number of origins was 14.01 for cotton and 13.01 for man-made fiber. Women's/girls' knit categories had effective numbers of 8.90 and 8.98, respectively. These category-level differences do not establish the availability of any specific polo, performance shirt or T-shirt.

The non-knit shirt categories provide another contrast. India led women's/girls' cotton non-knit shirts with 35.95%, while Bangladesh led men's/boys' cotton non-knit shirts with 26.74%. Vietnam led both man-made-fiber non-knit categories. A buyer who assumes that the largest origin in total apparel will also lead every shirt category would miss these distinctions.

The twelve-category comparison also shows why a concentration statistic should travel with a product definition. The highest HHI in this fixed frame was 1,895.58 for category 648, while the lowest was 713.63 for category 338. Those figures describe the observed category distributions. They should not be detached from the frame and presented as the extremes across all U.S. clothing imports.

Use the category table to sharpen a supplier question

A sourcing manager considering a new uniform range can use the table as a starting point for a more specific inquiry:

1. Identify the physical garment and its fiber composition.
2. Ask a qualified classifier or supplier which current commodity code applies; do not classify a product from its marketing name alone.
3. Compare the applicable source distribution with the garment's actual production origin.
4. Investigate factory capacity, material inputs and alternative specifications directly.
5. Record which parts of the analysis are national context and which are verified facts about the proposed order.

The table can help select questions. It cannot answer the factory-level questions on the buyer's behalf.

5. Which origins gained share, and which gained value?

Share and growth answer different questions. A share is relative to the whole. A growth rate compares an origin with its own earlier value. A contribution to total growth compares the origin's dollar change with the earlier World total. Keeping all three separate prevents a common reporting error: calling a relative gain an absolute expansion.

Enlarge figure ↗SVGPNGCSV

Read the data behind this figure (10 rows)

Share gains are not necessarily value growth: equivalent figure data

category_id	country	rank_2026	dollars_2025	dollars_2026	change_dollars	change_pct	share_2025_pct	share_2026_pct	share_change_pp	contribution_to_world_growth_pp	observed_date
1	China	3	6913609128	4555672975	-2357936153	-34.105719738340404	15.157992158777281	10.889676750999994	-4.268315407777287	-5.169742323632194	2026-09-14
1	India	6	3299645190	2453382879	-846262311	-25.64706998087876	7.234426330844079	5.864456611648629	-1.36996971919545	-1.8554183837867007	2026-09-14
1	Mexico	7	1534912085	1386247192	-148664893	-9.685564043233134	3.3652734654342598	3.3136232343063265	-0.05165023112793321	-0.32594571672456624	2026-09-14
1	Jordan	10	1167201910	1104083505	-63118405	-5.407668070042826	2.559074004897933	2.6391517875712065	0.08007778267327348	-0.1383862278516317	2026-09-14
1	Pakistan	8	1341865194	1268895467	-72969727	-5.437932761522988	2.942020833432942	3.0331109239550234	0.09109009052208128	-0.1599851147520816	2026-09-14
1	Bangladesh	2	4969379221	4658856821	-310522400	-6.2487161110138185	10.895295043632203	11.136322797595332	0.24102775396312914	-0.6808160567339355	2026-09-14
1	Italy	9	1159093974	1214610410	55516436	4.789640637023966	2.5412974676311495	2.903350353698198	0.36205288606704844	0.12171901621732249	2026-09-14
1	Indonesia	4	2660961334	2743329320	82367986	3.095422129873008	5.834120831654505	6.557531605161063	0.7234107735065578	0.1805906673065647	2026-09-14
1	Cambodia	5	2363611510	2621126266	257514756	10.894969622144037	5.182185465168191	6.265422895131236	1.0832374299630452	0.5645975321932382	2026-09-14
1	Vietnam	1	9429539296	9365565925	-63973371	-0.6784358067963875	20.674134169774582	22.387029550432175	1.7128953806575922	-0.1402607289528778	2026-09-14

Vietnam illustrates the distinction. Its January-July value decreased by about $64.0 million, or 0.68%, while its share increased 1.71 percentage points. Bangladesh's value decreased 6.25%, yet its share increased 0.24 points. Their declines were smaller than the decrease in the overall category-1 total.

Cambodia and Indonesia increased in both dimensions in this extraction. Cambodia's value increased 10.89%, and its share rose 1.08 points. Indonesia's value increased 3.10%, and its share rose 0.72 points. These are changes in customs values during the measured periods; they do not establish changes in individual factories' output or profitability.

China's value decreased by $2.358 billion, accounting arithmetically for 5.17 percentage points of the overall 8.28% decline. Its share decreased 4.27 percentage points. India's value decrease contributed another 1.86 points to the aggregate decline. These contributions describe the decomposition of the observed total. They do not establish a causal chain from a policy, order cancellation or production decision.

Why a trade shift is not automatically a factory move

An imported product's reported origin is not a record of the location of every upstream process. A finished garment can involve yarn, fabric, trims and services associated with multiple places. The current tables do not identify those upstream relationships, a buyer's ultimate ownership structure, or whether one supplier replaced another. Census's statistical guidance explains the basis of reported import origin; this study does not attempt an independent origin determination for particular products. [3]

The decline in one origin and increase in another could motivate further investigation, but they cannot demonstrate relocation by themselves. To document a factory move, an investigation would need identifiable firms, facilities, activities and dates. To document a sourcing switch, it would need comparable order or procurement records. Those are different studies from the aggregate analysis published here.

Similarly, a decrease in imported apparel value is not proof of U.S. reshoring. Domestic output, employment, inventories and demand would need to be examined on compatible bases. Arklavo's separate apparel-manufacturing report considers domestic industry data; combining those figures requires careful attention to what each series counts.

6. How the annual figures fit the 2026 comparison

Calendar-year 2024 category-1 imports were $79.225 billion; calendar-year 2025 imports were $77.612 billion in the current export. These full-year values are context for the trade series. They are not the denominator for a seven-month 2026 growth rate. [1]

Enlarge figure ↗SVGPNGCSV

Read the data behind this figure (4 rows)

Keep full years and matched seven-month periods separate: equivalent figure data

period	world_dollars	time_basis	scope	observed_date
calendar_2024	79225276317	full calendar year	OTEXA category 1	2026-09-14
calendar_2025	77611516848	full calendar year	OTEXA category 1	2026-09-14
jan_jul_2025	45610322631	January-July	OTEXA category 1	2026-09-14
jan_jul_2026	41834786093	January-July	OTEXA category 1	2026-09-14

The annual lead origin changed from China in 2024 to Vietnam in 2025. The top-five annual share decreased from 60.31% to 58.32%. The matched January-July comparison then shows a further decrease from 59.80% in 2025 to 57.24% in 2026. Each comparison stands on its own time basis.

Seasonal patterns and the timing of entries are reasons to retain the matched-period comparison. Dividing a full year by twelve assumes that each month is comparable. Multiplying seven months by twelve-sevenths assumes the remaining months will follow the same average. Neither assumption is used in this report.

A source-version note for researchers

The live export's January-July 2025 total is $45,610,322,631. The September 5, 2026 Daily Star summary reports $41.83 billion for the current period and an 8.65% decline. The current official export produces 8.28%. This report preserves the downloaded figures and extraction timestamp rather than adjusting its arithmetic to match that secondary summary. The available evidence does not establish the cause of the discrepancy. [1,7]

This is a material reason to publish the underlying data. A reader can inspect the exact numerator and denominator, repeat the calculation and compare them with another dated release. If a future official export changes either number, it should be treated as a new data version, with a recorded update to the table, figures and narrative. Quietly replacing the denominator would make the report harder to cite and audit.

The same principle applies to catalog and certification research elsewhere in this series. A date-stamped observation is a record of what was accessible at that time. It is not a promise that a website, catalog or dataset will remain unchanged.

7. Turn the benchmark into a purchasing decision

The report is most useful when it helps a reader move from a broad market question to evidence about a specific garment order. National data can flag the category and origin relationships worth investigating. The buyer's actual exposure depends on the order and supply arrangement.

Build a minimum sourcing record

The record should distinguish the seller, the manufacturing facility, the reported garment origin, the decoration location and the upstream material sources. They may differ. Buying from a U.S. distributor or receiving domestic decoration does not by itself establish where the blank garment was manufactured.

A useful sourcing record also states the confidence level for each field. A supplier statement, a purchase-order commitment, a current stock confirmation and an inspected supporting document provide different kinds of evidence. Record what was actually supplied rather than filling gaps with a country's general reputation.

A five-step exposure review

1. Define the operational requirement. List the garment, size and color combinations, required quantities, destination and needed-in-hand date. Include replacement demand if the uniform must remain available beyond the first order.
2. Confirm the present supply route. Obtain the blank product's origin and the production or stock location relevant to this order. A brand-wide factory list is not an order-specific answer.
3. Find shared dependencies. Ask whether supposedly separate products or vendors rely on the same fabric source, finishing process, warehouse or transport route. Geographic variety alone may leave these dependencies unchanged.
4. Test substitutions against the specification. Compare fabric, color, fit, decoration result and continuity. A substitute that fails the role or the wearer's needs does not solve the original purchasing problem.
5. Assign a dated action. Record who will reconfirm stock, approve samples, review documents or authorize the alternative. Retain the evidence used for that decision.

This process does not require choosing the origin with the smallest national share. A small trade flow can reflect a narrow specialty, a small market or other circumstances that the aggregate data cannot identify. The appropriate choice depends on the specification, evidence and buyer's operational priorities.

8. A transparent exposure scenario

An illustrative scenario can make the distinction between a market benchmark and a company decision concrete. Suppose a buyer assigns a fraction of its order to one supply route and then assumes that part of that route becomes unavailable. With no inventory buffer, substitution or recovery, the directly affected order share is:

Affected order share = assigned route share × assumed unavailable fraction.

This identity is a planning calculation. It is not a prediction of disruption, an estimated probability, a loss forecast or a statement about any origin's reliability. It deliberately omits timing, safety stock, recovery and correlated failures so that the assumption is visible.

The middle row uses a share close to the observed national leader purely to demonstrate the arithmetic. It does not imply that an individual buyer has that exposure. The user should replace it with the actual assigned order share and an explicitly chosen assumption. The downloadable scenario sheet contains these inputs and formulas.

Explore an assumed route exposure

This planning identity uses your assumptions. It does not estimate the likelihood of a disruption.

Printed example: a 40% route share × 50% assumed unavailability ÷ 100 = 20% of the order. No inventory buffer, substitution or recovery is included.

To make the scenario operational, add evidence about stock already held, the time needed to approve an alternative, and the quantity that could be replaced without changing the garment specification. A resulting decision may concern a second approved style, a color substitution, a different order schedule or a larger buffer. This report does not recommend a fixed buffer size or a universal number of countries.

9. How this analysis relates to established industry reports

The 2026 USFIA Fashion Industry Benchmarking Study examines participating companies' sourcing experiences and plans. Its survey perspective can explain the questions businesses are asking and the strategies they report considering. Our trade analysis measures recorded flows, which is a different type of evidence. An intention to diversify and a change in country-level import concentration need not move together. [5]

The U.S. International Trade Commission's 2024 apparel export-competitiveness report examines foreign suppliers and the factors shaping their competitiveness. Its country-level research provides a broader analytical frame than a single table of import values. The present report contributes a current, narrow and reproducible comparison across a fixed set of categories; it does not replace the Commission's investigation or claim equivalent research breadth. [6]

Our contribution is the connection between matched 2026 trade periods, category-specific concentration, downloadable country-level arithmetic and an order-level evidence template. A reader can trace a headline through its denominator to the data row and then identify which additional facts are needed for a purchasing decision.

The main comparative weakness is also clear. We have not interviewed factories, surveyed a representative set of buyers, measured lead times or independently examined corporate supply chains. The report supplies a verifiable analytical layer. Its practical templates still require validation by actual users and their suppliers.

10. Methods, reconciliation and reproducibility

Acquisition and the fixed frame

The study protocol was written after a small read-only access pilot and before the full export or concentration analysis. The pilot established that the official explorer displayed July 2026 data and permitted a summarized export. It did not select categories according to their results. The thirteen category codes were fixed as 1, 338, 339, 340, 341, 347, 348, 638, 639, 640, 641, 647 and 648.

The export was obtained through the source's visible controls: Annual Data, Category selection, those thirteen categories, Dollars, with Country, Chapter and HTS otherwise unfiltered. The displayed row count was 2,746, matching the downloaded CSV. Four period columns were converted into 10,984 category-origin-period observations. The rolling-year columns were retained in the private source export but not used in this analysis.

Rows for World and eleven overlapping country groups were separated from individual reporting-area rows. The official selector's country/group classification was retained alongside the export. Group observations were not added to individual origins or to each other. The analysis uses the exported origin name as the unique reporting key because some selector labels have multiple historical codes.

Two legacy export labels, Yugoslavia and Yugoslavia [2], were absent from the current selector under those exact names. They were retained as explicitly unresolved legacy rows; every value in the four analyzed periods was zero. The process asserts that condition before continuing. The selector also contained a conflicting regional classification for Grenada. The raw metadata and joined code values remain in the classification file; no continental aggregation or geopolitical conclusion uses that metadata.

Denominators and formulas

For each category-period, all individual reporting-area dollar values were summed and compared with the corresponding World value. All 52 category-period reconciliations matched exactly, to the dollar. No result depends on rescaling an incomplete country list to 100%.

All calculations use unrounded integer dollar observations. Rounding is applied only to displayed results. Country rankings use descending 2026 value, with a deterministic alphabetical tie-break. A positive-origin count includes only rows with a value above zero. A zero base has no calculated growth rate. Missing numeric values would stop the analysis rather than being converted silently into zeros.

The effective number is a restatement of HHI, not independent corroboration of it. CR1, CR3, CR5, HHI and positive-origin counts describe different features of the same dataset. Their agreement or disagreement should not be described as multiple independent studies.

What another researcher can reproduce

The reader download includes the factual observations, category frame, country classifications, concentration measures, origin shares, matched-period changes, reconciliation table, scenario inputs and a standard-library Python script. Running python reproduce.py --check in the trade-report directory recalculates the findings and compares them with the stored result file. Running python reproduce.py rebuilds the derived tables.

The source-to-result chain is: official export → classified observations → reconciled denominators → shares and concentration → figure data → manuscript. Each figure supplies its own CSV and both SVG and PNG versions. The source URL and extraction date allow a researcher to obtain a later version and repeat the analysis, with a new version note rather than overwriting this observation silently.

Limitations that affect interpretation

• Coverage: OTEXA's defined apparel categories do not equal every possible retail definition of clothing, and the twelve detailed categories are not a uniform-market census.
• Measurement: values are nominal general imports. They are not prices, quantities, final consumption, domestic value added or company revenues.
• Granularity: reported origins are not factories, owners, certified supply chains or interchangeable production options.
• Time: seven months of 2026 do not establish the full-year outcome. Current data can be revised.
• Causality: observed changes do not identify the effect of tariffs, demand, inventory policy or relocation.
• Metadata: historical labels and selector inconsistencies are preserved; unsupported region-level findings are avoided.
• Application: the purchasing workflow is an analytical aid, not a representative user-tested intervention or a supplier endorsement.

Arklavo sells decorated apparel and is the publisher of this report. That commercial interest is disclosed because the study may inform uniform purchasing. No supplier paid for a position in the tables. The results do not demonstrate that Arklavo or any named company has a more resilient supply chain.

11. Questions buyers and researchers ask

Which country supplies the most apparel to the United States in 2026?

Vietnam was the largest reported origin by value in OTEXA category 1 for January-July 2026 in the September 14 extraction, supplying $9.366 billion, or 22.39%. This answer is limited to the defined category, period and measure. It is not a full-year ranking or a count of garments.

Are U.S. apparel imports becoming less concentrated?

Across the matched January-July periods, category-1 CR3, CR5 and HHI declined, while the largest origin's share increased. The answer depends on the concentration measure and category. A statement about the overall distribution should identify both, rather than substituting a single country share for the whole distribution.

Does a fall in Chinese apparel imports mean the work moved to the United States?

No such conclusion follows from this dataset. It records imports by reported origin. Establishing a transfer to domestic production would require compatible domestic production evidence and, for a claim about a particular move, identifiable firm and facility records.

Is import dependence the same as import concentration?

No. Imports as a share of consumption require a domestic-demand or supply denominator. Concentration describes how a given set of imports is distributed across origins. This report measures the latter. A country can have a diverse set of import origins and still rely heavily on imports overall.

How many countries should a uniform buyer source from?

This analysis does not identify an optimal number. A country count does not measure shared mills, approved capacity, product equivalence, inventory or delivery performance. Establish the order's requirements and dependencies before setting a diversification target.

Can I use the figures in a presentation or article?

The original Arklavo figures and factual data tables are available for reuse with attribution and their source-period and scope notes retained. Cite the specific report edition and figure. Third-party source documents remain subject to their publishers' terms; the reader package links to them rather than republishing their full reports.

12. Conclusion: what the latest 2026 evidence changes

The 2026 Import Concentration Report shows a sourcing distribution that is changing in more than one direction. Vietnam's share of the defined apparel total increased, while the top-three and top-five shares and the all-origin concentration index decreased. Individual categories were materially different from the national aggregate: the leading-origin share spanned 13.65% to 39.24% in the fixed twelve-category comparison.

The most useful response is more precise investigation. Match the product to the relevant category, preserve the period and denominator, and then verify the actual order's origin, facilities, stock and alternatives. Those steps turn a national benchmark into a set of answerable procurement questions.

This edition preserves the latest official July 2026 observations available in the explorer on September 14, 2026, alongside the older periods needed for comparison. Its downloadable arithmetic makes the findings inspectable. Future releases can update the results; they cannot supply factory-level facts that the original data do not contain.

For the domestic industry context behind a U.S. sourcing decision, see our Apparel Manufacturing in the USA: 2026 Report.

References and data downloads

1. Office of Textiles and Apparel. U.S. Textile and Apparel Imports by Category or Country, linked Annual Data explorer. Official summarized CSV exported September 14, 2026; display period July 2026. Original calculations: Arklavo Research.
2. Office of Textiles and Apparel. Imports Headnotes. Scope, general imports, aggregate groups and category exclusions; checked September 14, 2026. The distinction between general imports and imports for consumption follows the Census definition in reference 3.
3. U.S. Census Bureau. Guide to the U.S. International Trade Statistical Program, section 2. Coverage, statistical month, customs valuation and country-of-origin definitions; checked September 14, 2026.
4. Office of Textiles and Apparel. Trade data directory and the Annual Data explorer's current category and country selectors. Category descriptions and classification metadata captured September 14, 2026.
5. U.S. Fashion Industry Association. 2026 Fashion Industry Benchmarking Study. Used as a methodological comparator for company survey evidence, not as the source of the trade calculations.
6. U.S. International Trade Commission. Apparel: Export Competitiveness of Certain Foreign Suppliers to the United States, 2024. Used as a country-research comparator; not described as a 2026 study.
7. The Daily Star. Bangladesh overtakes China again in apparel exports to US, September 5, 2026. Consulted only to identify the published numerical discrepancy; the official export remains the analytical source.

Suggested citation: Arklavo Research. (2026, September 14). U.S. Apparel Supply Chains: The 2026 Import Concentration Report. Arklavo. https://arklavo.com/blogs/custom-apparel-guide/apparel-supply-chains

Research contact: Arklavo Research (info@arklavo.com). For a data correction, identify the category, period, source row and proposed correction so the result can be checked.