Tolerance Stack-Up: Right Measurements, Wrong Fit
Why garment measurements can pass tolerance while the fit is still small: how shrinkage, cutting and sewing deviations add up, and how to control the stack.

Answer first: Tolerance stack-up is what happens when several small deviations push in the same direction. Fabric shrinkage after washing, cutting variation, seam allowance taken slightly wide and stitching that pulls in a seam can each stay inside its own limit, yet together move a finished measurement outside what the customer experiences as the right size. A T-shirt can pass a chest tolerance at inspection and still fit small after the first wash, because inspection measures the garment as shipped, not as worn. The fix is to treat tolerances as a budget: decide the total deviation the fit can absorb, allocate it between fabric, cutting and sewing, and measure the biggest contributor, usually shrinkage, with a standard test such as ISO 5077 before the pattern is finalised.
This article explains the arithmetic with a worked example. The example figures are illustrative inputs for the calculation, not measurements from a real order. The ISO 5077 details were checked on 17 November 2026.
Where deviation comes from
Every finished measurement is the result of a chain. Each link adds its own variation.
| Stage | Source of deviation | Typical direction |
|---|---|---|
| Fabric | Residual shrinkage after washing; width and weight variation between lots | Mostly smaller after washing |
| Pattern | Shrinkage allowance set from the wrong test or the wrong fabric lot | Systematic, either way |
| Spreading | Fabric relaxed or under tension when laid | Tension usually gives smaller panels after relaxation |
| Cutting | Blade drift in high plies; marker accuracy | Random, either way |
| Sewing | Seam allowance variation; seam pull-in; rib attachment tension | Usually smaller |
| Finishing | Pressing, steaming, garment wash | Either way, depending on process |
| Measuring | Measuring method, garment laid flat or stretched | Random, either way |
The important distinction is between systematic deviations, which push all garments the same way, and random deviations, which vary from piece to piece. Systematic deviations add up directly; random ones partly cancel. Shrinkage is the classic systematic deviation.
Shrinkage: the largest link
How it is measured
ISO 5077:2007 specifies how to determine the dimensional change of fabrics, garments or other textile articles after washing and drying. From the published preview:
- The specimen is conditioned, measured, washed and dried using a procedure from ISO 6330, then conditioned and measured again. Marking and measuring follow ISO 3759.
- Dimensional change is calculated as (x_t − x_o) / x_o × 100, where x_o is the original and x_t the measurement after treatment.
- Results are expressed to the nearest 0.5%, with shrinkage shown as a minus sign and extension as a plus sign.
- For garments, the report gives the change at each measuring position.
Two details cause most disputes: the washing procedure chosen from ISO 6330 and the number of cycles. Agree both in writing, and test the bulk fabric, not only the sample fabric.
Why it stacks with everything else
If the pattern is cut with an allowance for 4% length shrinkage and the bulk lot shrinks 6%, every garment ends up 2% shorter than intended after washing. On a 72 cm body length, that is about 1.4 cm. If the length tolerance at inspection is ±1.5 cm, the shipped garment passes, but the washed garment may sit at the edge of tolerance before any cutting or sewing deviation is added. Our article on shrinkage allowance in T-shirt patterns explains how the allowance is set.
Cutting and sewing
Cutting
In a high lay, the top and bottom plies can differ slightly in size if the blade leans or the lay shifts. Marker accuracy matters too: a marker plotted at the wrong scale, or for the wrong shrinkage allowance, moves every panel. Finetex’s quality control page lists a cutting check of marker accuracy, ply count and panel dimensions on the first bundle of each size for this reason.
Sewing
Each seam consumes a seam allowance. A body with two side seams and an operator who takes 2 mm more than specified on each side loses 8 mm of chest circumference (2 mm × 2 seams × 2 layers of fabric per seam). Rib collars and cuffs attached under too much tension draw in the opening. Seam types and stitch classes are catalogued in ASTM D6193, the standard practice for stitches and seams, which is a useful reference when the specification names the seam construction.
Calculating the stack
Worked example
Take a T-shirt chest measured flat (half chest), with a specification of 52.0 cm and a tolerance of ±1.0 cm. These inputs are illustrative:
| Contributor | Type | Effect on half chest |
|---|---|---|
| Width shrinkage higher than the pattern allowance (1% of 52 cm) | Systematic | −0.5 cm |
| Spreading tension, panels relax smaller | Systematic | −0.2 cm |
| Seam allowance 2 mm wider per side seam | Systematic | −0.4 cm |
| Cutting variation | Random | ±0.3 cm |
| Measuring variation | Random | ±0.3 cm |
Two ways to add it up
- Worst case (arithmetic sum). Add all contributions in the same direction: −0.5 − 0.2 − 0.4 − 0.3 − 0.3 = −1.7 cm. The garment is outside a ±1.0 cm tolerance.
- Statistical (root sum of squares) for the random part. Systematic deviations still add directly: −1.1 cm. The random part combines as √(0.3² + 0.3²) ≈ ±0.42 cm. The likely range is −1.1 ± 0.42 cm, so −0.68 to −1.52 cm.
Either way, the systematic part alone (−1.1 cm) already exceeds the tolerance, even though no single contributor looked large. At the inspection table, before washing, the shrinkage term is not visible, so the garment might measure −0.6 cm and pass. After the customer washes it, it is about 1.1 cm smaller than the specification on the half chest, more than 2 cm in full circumference.
Reading the result
- Systematic deviations are the priority: they do not average out, and the inspection sample will not reveal shrinkage that happens after washing.
- Random deviations set the minimum tolerance you can realistically hold; asking for ±0.5 cm when measuring alone varies by ±0.3 cm creates disputes rather than quality.
Control points
1. Test shrinkage on the bulk lot
Run the agreed ISO 5077 procedure on the bulk fabric before cutting and compare it with the pattern’s allowance. Finetex’s incoming fabric inspection includes shrinkage checks on the lot. If the result differs, adjust the marker, not the tolerance.
2. State the measuring method
Define each point of measure, whether the garment is laid flat or measured in circumference, and how much tension is allowed. Our article on garment measurement points for knit tops lists the common ones.
3. Budget the tolerance
Allocate the total allowed deviation between fabric, cutting and sewing. For a key point such as chest, a practical budget might reserve the largest share for shrinkage, then seam allowance, then cutting. Write it into the tech pack.
A budget can be as simple as one table per fit-critical point. The split below is a format to copy, with illustrative numbers for a ±1.0 cm half-chest tolerance:
| Link | Allowed share | How it is controlled |
|---|---|---|
| Residual shrinkage after washing | ±0.4 cm | Bulk fabric test, marker adjusted to the result |
| Seam allowance and pull-in | ±0.3 cm | Seam allowance on the pattern, operator check at first output |
| Cutting | ±0.2 cm | First-bundle panel measurement per size |
| Measuring | ±0.1 cm | One written method, same tape and table |
If any link needs more than its share, the budget shows which other link must be tightened, or whether the tolerance itself is unrealistic for the fabric.
4. Check the first bundle and the first output
Measure the first panels from each size at cutting and the first garments from the line. If the systematic direction shows up there, correct the operation before the run continues. Our article on the first-output check describes the routine.
5. Measure after washing, not only as shipped
For fit-critical styles, add a measurement of washed garments to the pre-production approval. Our guide to underwear measurement points and tolerances applies the same logic to underwear, where elastic and fabric recovery add another link.
6. Track drift per batch
Finetex measures ten garments at the start and end of each batch against the size chart, as stated on our quality control page. A consistent drift in one direction, even inside tolerance, is a systematic deviation to correct.
When the customer says “it fits small”
- Measure the returned garment flat, then wash an unworn garment from the same lot and measure it again.
- Compare both with the specification and the pattern’s shrinkage allowance.
- If washed garments are consistently small, the cause is the shrinkage link; adjust the allowance or the fabric finish.
- If unwashed garments are already at the lower edge, check seam allowances and rib attachment.
- Record the finding and the correction in the order file before the reorder. Our article on tolerance disputes explains how to resolve these with the factory.
Fact, advice and uncertainty
Fact: ISO 5077 measures dimensional change after washing and drying, expressed to the nearest 0.5% per measuring position; systematic deviations add directly, while random ones partly cancel.
Advice: budget tolerance across fabric, cutting and sewing; test shrinkage on the bulk lot; and measure washed garments for fit-critical styles.
Uncertain: how much each link contributes for your product. That depends on fabric, construction and factory practice, and should be measured rather than assumed; the figures in our example are illustrative only.
Related reading
- Garment Measurement Points for Knit Tops: The 12 Numbers Buyers Argue About
- Grading Rules for Knit Tops: Where Size Sets Fail
- Vanity Sizing: Adjust the Chart or the Pattern?
- A Fit Session Protocol for Knit Tops: 12 Wearers, 5 Measurements
Frequently asked questions
What is tolerance stack-up in garments?
It is the combined effect of small deviations in fabric shrinkage, cutting and sewing. Each can be inside its own limit while the total moves the fit outside what customers accept.
Why does a garment pass inspection but fit small after washing?
Inspection usually measures garments as shipped. If the fabric shrinks more than the pattern allowed for, the garment becomes smaller only after the first wash.
Should I tighten tolerances to fix fit problems?
Usually not. Find and correct the systematic deviation, often shrinkage or seam allowance. Tolerances tighter than measuring variation cause disputes without improving fit.
Which test measures garment shrinkage?
ISO 5077 is widely used, with washing and drying procedures from ISO 6330 and marking and measuring per ISO 3759. Agree the procedure and number of cycles in writing.
Sources
- ISO 5077:2007 preview: Determination of dimensional change in washing and drying
- ASTM D6193: Standard Practice for Stitches and Seams
Send us a garment that fits small
Post us a sample with your size chart and fabric details. We will measure it, wash-test the fabric and tell you which link in the stack is causing the problem before we quote a corrected specification.
Request a quote or read about our quality control gates and T-shirt fit and fabric guide.
More factory-floor notes for brand owners

Importing Apparel into Japan: JIS Care Symbols and Quality Labelling
Japan's Household Goods Quality Labeling Act for apparel: fibre composition terms, JIS L 0001:2024 care symbols, labeller details…
Read more →
Apparel Manufacturing Glossary: 120 Terms
120 garment manufacturing terms in English and Chinese with one-line explanations, grouped in 12 topics, for buyers working…
Read more →
Subcontracting: How to Know If Your Order Is Outsourced
Six signs your apparel order has been subcontracted, the contract clauses that control it, and how to check…
Read more →
Gym, Studio and Team Kit: Low-Volume Custom Orders
How gyms, studios and teams can order custom kit in small runs: one blank body, names and numbers,…
Read more →
Raglan and Baseball Tees: Construction and Decoration
How raglan and baseball T-shirts are built: raglan seams, collar attachment, print placement across seams, two-colour shade control…
Read more →
The Cost of Poor Quality: Rework, Chargebacks and Air Freight
Calculate the full cost of a 2% defect incident in an apparel order: rework, re-inspection, air freight, chargebacks…
Read more →Request a sample & pricing information
Send us your tech pack, target price or a reference sample — we reply with a quotation, a realistic lead time and a fabric recommendation within one working day.
- Free sampling plan & cost breakdown
- MOQ from 300 pcs per style
- Reply within 1 working day