ROUND 1 · THE WHITE TEE BASELINE
In the first round of the experiment, I tested T-shirts from ARKET, ASKET, H&M, Levi’s, Loom, Samsøe Samsøe, SANVT and UNIQLO.
The prices ranged from €6.99 to €50. I wore each T-shirt for between 125 and 129 hours and washed it ten times, following the same washing and drying procedure. I measured the garments before testing, after the first wash and after the tenth wash. I also examined their construction, comfort, appearance and durability, and contacted the brands to investigate how much information they could provide about their supply chains.
The differences between the eight T-shirts were striking.
What ten washes revealed
One of the most important findings concerns shrinkage.
H&M’s T-shirt lost 5.6 cm2.2 inches in length, shrinking from 73.0 to 67.4 cm28.7 to 26.5 inches. UNIQLO’s Supima T-shirt performed almost as badly, losing 5.2 cm2.0 inches. Its sleeves also became noticeably shorter. As one would expect, these changes substantially affected how the T-shirts fitted: whereas their fit was initially unproblematic, they became somewhat too short.
At the other end of the spectrum, ASKET and SANVT showed limited dimensional change. Loom also remained relatively stable, although its shoulders narrowed a bit and slight twisting developed.
SANVT, however, developed pronounced twisting despite its relatively stable measurements. Levi’s also showed substantial twisting, which became apparent after the first wash.
ARKET, meanwhile, showed no visible twisting even after ten washes, although it shrank noticeably across the chest, shoulders and hem. Samsøe Samsøe showed moderate shrinkage, particularly at the shoulders and sleeves, along with noticeable twisting.
These differences show why looking at dimensional changes alone is not enough to assess durability.
How the T-shirts held up
| T-shirt ↗ | Length change | Twisting | Pilling |
|---|---|---|---|
| ARKET | −1.5 cm−0.6 in | 0None | 0None |
| ASKET | −0.9 cm−0.4 in | 2Noticeable | 0None |
| H&M | −5.6 cm−2.2 in | 3Pronounced | 0None |
| Levi’s | −0.6 cm−0.2 in | 3Pronounced | 0None |
| Loom | −1.0 cm−0.4 in | 1Slight | 0None |
| Samsøe Samsøe | −1.7 cm−0.7 in | 2Noticeable | 0None |
| SANVT | −1.3 cm−0.5 in | 3Pronounced | 0None |
| UNIQLO | −5.2 cm−2.0 in | 2Noticeable | 3Widespread |
Length change compares new condition with after ten washes. Twisting and pilling use the WTL structural-change scale (0 = none; 1 = slight; 2 = noticeable; 3 = pronounced/widespread). Click a T-shirt to read the individual review.
Comfort, construction and appearance
Comfort was generally less of a problem than durability. The T-shirts generally remained pleasant enough to wear throughout the test, even when their appearance or dimensions had changed noticeably.
ASKET and Loom both felt relatively substantial, and neither developed major problems with fit. Their fabrics did, however, lose some of their initial softness. This was especially clear with Loom, where I could compare the tested garment with an unused example of the same T-shirt. ARKET also lost some of the softness it had when new, but remained comfortable after ten washes.
Construction quality varied considerably. H&M had several untidy seams and protruding threads, some of which were already visible before washing. Samsøe Samsøe also showed relatively poor finishing, with loose threads and increasingly untidy stitching around the collar. Levi’s had similar issues with loose threads along internal seams.
One of the more surprising results concerned pilling. Seven of the eight T-shirts showed no visible pilling after ten washes. UNIQLO was the clear exception, developing widespread pilling across the fabric surface. Combined with its substantial shrinkage and the twisting already visible when new, this left the T-shirt looking considerably worse than it had at the start.
Was spending more worth it?
The two strongest performers were ASKET’s The T-Shirt (€50) and Loom’s Le T-shirt Classique (€25).
ASKET showed particularly good dimensional stability, neat construction and consistent comfort. Loom was similarly comfortable and generally well constructed, although it showed somewhat more shrinkage and lost some of its initial softness. Still, at half the price of ASKET, I’m impressed by Loom’s performance.
In general, the relationship between price and performance was far from straightforward. ARKET (€20) performed well, while considerably more expensive T-shirts from Levi’s (€44) and SANVT (€45) developed pronounced twisting. SANVT otherwise showed good dimensional stability.
H&M (€6.99) was by far the cheapest T-shirt in the round. However, its substantial shrinkage, pronounced twisting and untidy construction make it difficult to recommend, even as a budget option.
Overall, some relatively affordable alternatives competed favourably with much more expensive ones, although the cheapest T-shirt was the poorest performer together with UNIQLO (€19.90).
Price and overall assessment
| T-shirt ↗ | Regular price | WTL Rating |
|---|---|---|
| Loom | €25 | |
| ASKET | €50 | |
| ARKET | €20 | |
| SANVT | €45 | |
| Samsøe Samsøe | €33.50 | |
| Levi’s | €44 | |
| H&M | €6.99 | |
| UNIQLO | €19.90 |
Regular prices recorded for Round 1. The WTL Rating is a whole-number assessment of each tested garment, not a calculated average of the measurements. The individual reviews explain each rating.
The transparency problem
Alongside the wear and wash tests, I contacted all eight brands to investigate how much information they would be willing or able to provide about the production of their T-shirts.
I focused on five stages of production: cotton farming, spinning, knitting, dyeing and finishing, and garment manufacturing.
Here, too, the differences were considerable.
Loom, ASKET and SANVT provided the most detailed accounts of their supply chains, identifying all five stages. Loom and ASKET named specific suppliers, while SANVT identified the locations of its production processes.
However, identifying production stages was not the same as documenting them. None of these three brands provided sufficient evidence for me to independently verify the complete supply chain of the particular T-shirt I tested.
Loom supplied GOTS-related documentation, but this did not establish the supply chain of the finished garment. ASKET was relatively open about the limitations of its information, including its inability to verify the cotton origin for the specific product.
For example, Loom named the spinning mill, knitting mill, dyehouse and garment factory. What I did not receive was a set of product- or batch-linked records – such as relevant GOTS Transaction Certificates and corresponding production records – connecting those suppliers to the particular T-shirt I purchased. Knowing the names of the facilities was therefore different from being able to verify that the particular garment passed through them.
Other brands provided considerably less information. ARKET identified its garment supplier in Bangladesh, but not the preceding production stages. H&M and Levi’s disclosed the countries of manufacture without identifying the factories responsible for the tested T-shirts.
Samsøe Samsøe presented an interesting case. According to the brand, its T-shirt was GOTS-certified as a finished product, yet it could not identify the garment factory or provide details of the earlier production stages. UNIQLO supplied some information about its Supima cotton sourcing, but the intermediate production stages could not be linked to the T-shirt I tested.
Overall, sustainability reports, certifications and supplier lists do not necessarily make it possible to trace an individual product. Even when brands were willing to answer my questions, independently verifying their product-specific claims proved remarkably difficult.
Supply-chain information
| Brand ↗ | Stages identified | Brand transparency |
|---|---|---|
| Loom | 5 of 5 | High |
| ASKET | 5 of 5 | Medium |
| SANVT | 5 of 5 | High |
| ARKET | 1 of 5 | Low |
| Samsøe Samsøe | 0 of 5 | Medium |
| H&M | 0 of 5 | Low |
| Levi’s | 0 of 5 | Low |
| UNIQLO | Pending | Medium |
Five stages: cotton farming, spinning, knitting, dyeing/finishing and garment manufacturing. Identified does not mean independently verified for this garment or its production batch. UNIQLO’s stage assessment remained pending. Brand Transparency concerns disclosure, not environmental performance.
What I take away from Round 1
The most striking finding is how differently eight relatively ordinary white cotton T-shirts behaved under the same test conditions. After ten washes, some retained their dimensions but became twisted. Others avoided twisting but shrank considerably. Most resisted pilling, but one did not.
Two of the T-shirts stood out: ASKET and Loom. Neither was perfect, but both remained comfortable, generally well constructed and relatively stable throughout the test. Loom was particularly impressive given its price.
ARKET also performed well for a relatively inexpensive T-shirt, while SANVT demonstrated that good dimensional stability did not necessarily prevent serious twisting.
H&M and UNIQLO were the biggest disappointments. Both became substantially shorter, while H&M developed pronounced twisting and UNIQLO showed widespread pilling.
The investigation into supply-chain transparency revealed similarly large differences, although even the most informative brands fell short of providing fully verifiable product-specific documentation.
These are the results of testing eight individual garments, not evidence that every T-shirt from the same brands will perform identically. Ten washes should also only be considered an early indication of durability, rather than a complete picture of long-term performance.
Nevertheless, the first round provided a useful baseline for the experiment. It showed how much there is to learn from treating an apparently simple product — the ordinary white T-shirt — as something worth examining closely.