Apparel Size Chart & Grading: OEM Fit Guide
To build an apparel size chart that survives grading, land six modules — base size, POM points, grading rules, graded tolerances, full-size fit samples, and market-size alignment — because a chart with only a base size and no grade rule is half a spec and the source of "your XL is not my XL."
- Grade by fabric: woven adds ~4 cm chest per size; 4-way stretch adds only ~2 cm, or large sizes run tighter — apply the wrong rule and fit drifts.
- Tolerances: ±1.0 cm for structured parts, ±1.5 cm for stretch; a flat ±1 cm across all points can let neighbouring sizes differ by 2 cm.
- Market alignment: the same "M" differs 4–6 cm in chest across US, EU, UK, and Asian cuts — pick the target market before you grade.
- Fit samples: validate a full-size set, not just the base; one real case measured a finished XL 1.5 cm under its graded expectation.
Bottom line: A size chart is only complete with six modules and a grade rule matched to fabric and market — nail the base size first, then grade, or grading only magnifies the error.
Apparel Size Chart & Grading: OEM Fit Guide for Sportswear
"Send me your size chart." It is the line most buyers write into a tech pack and then stop thinking about — and the line that decides whether a shipment fits. Many brands treat the size chart as a formality: one base size, a few measurements, and the rest left to the factory. It is not. A size chart without a grading spec is half a specification, and the missing half is exactly where "your XL is not my XL" complaints are born.
Last year a European yoga brand came to us with 2,000 returned tops. Their previous factory — a competitor — had shipped them on time, and the sealed M size had been signed off by both sides without a word. When the returns came in, the complaint was consistent: the XL ran tight. We measured a finished XL against the sealed M and found the chest came in 1.5 cm smaller than the graded-up expectation. The root cause was not a sewing error. The tech pack had given only the M size. There was no grade rule — no statement of how much each size should grow or shrink. The competitor graded on its own default, and its default XL was tighter than the brand's mental XL.
Here is the counterintuitive truth: most "wrong size in bulk" complaints do not start on the sewing line. They start at the spec, where the size chart never said how much neighbouring sizes should differ. The buyer gave a base size but no grading increment; gave a fabric composition but no rule for how a stretch fabric should shrink the grade. The factory can only grade by its own habit — and its habit is not your fit.
This is written for sourcing managers, product developers, and founders of sportswear brands. We are a sportswear factory supplier running OEM and ODM production in Dongguan since 2014, across two facilities totalling 20,000㎡ with a monthly capacity of 500,000 pieces. What follows — how to build a size chart that survives grading, how to write a grade rule that holds, and how to catch fit problems before bulk — comes from working through 1,200+ brand orders, not from a textbook. And because fit and inspection are two halves of the same shipment, our AQL inspection guide covers the quality-control half of the same problem.
The Five Ways Brands Get Size Charts Wrong
- Base size only, no grade rule. The chart lists M and leaves L/XL/XXL to the factory default. Your M in another system may be their L — the whole size run drifts.
- Stretch fabric graded on a woven logic. Brands treat a base layer like a woven shirt: +4 cm chest per size. But 4-way stretch recovers, so a woven-grade XL ends up too tight to wear. We have seen this too many times.
- One flat tolerance for everything. A chart that writes ±1 cm on all 12 points of measure can, in the worst case, let neighbouring sizes differ by 2 cm on a point. Flat tolerance stacks error.
- Mixed market systems. The same "M" differs 4–6 cm in chest between US, EU, UK, and Asian cuts. A tech pack that does not name the system gets flagged by end consumers as "off-size" the moment it hits shelf.
- Skipping the fit sample. Signing one base size and going straight to bulk drops the only pre-bulk gate that catches fit. No other size in the run was ever worn or tested.
Why these five traps turn straight into money: a size run that drifts by a size is not "fix a few pieces." It means returns at retail, rework on the size set, and a trust cost that no unit price ever shows. A grading spec written in the wrong place never shows on the quote. It shows after sales — in the return rate.
What a Workable Size Chart + Grading Spec Must Carry
A spec that can go straight to the cutting room and not drift in bulk must state six blocks. Miss any one and the consistency of your size run rests on luck, not on the document.
1. Set the Base Size
Use your best-selling size as the master (base size). Do not use the smallest or the largest size as the base. The base is the anchor every other size is graded up or down from, so the closer the base sits to your main sales size, the more accurate the whole run. We default to M for men's and S/M for women's as the base. The reason is mechanical: grading is "add or subtract from the base," so if the base itself is right, the rest inherit that rightness. A base that is your hero size also means the size you sell most is the one measured and fit-tested most carefully — exactly where an error would hurt most.
2. Write the POM (Points of Measure)
Every dimension must state where it is measured: chest is taken how many centimetres below the armpit, body length from the centre-back neck point to the hem, sleeve length from which shoulder point to where. On the same garment, a 2 cm shift in the point of measure can read as half a size. We list POM per garment type and keep the list consistent between the tech pack, the factory floor, and your QC — because if POM is not written, the sewing line and your inspector will never reconcile the numbers. Common points we always pin down: chest (1 cm below armhole), waist (at the narrowest), hip (fullest), body length (CB neck to hem), sleeve length (shoulder point to cuff), shoulder width, neck opening, and hem width. Each gets one unambiguous reference line.
3. Give the Grade Rule
This is the most common omission, and the most expensive. The spec must state how much each point changes per size step. As a benchmark, woven men's sportswear commonly uses +4 cm chest per size, +2 cm body length, +1.5 cm sleeve. But knit and stretch fabrics must be cut roughly in half or less — for 4-way stretch we recommend starting at +2 cm chest per size, otherwise the larger sizes get tighter as they scale, never looser. The grade rule is the sentence that says "this is what your XL looks like." Without it, you have handed the definition of your fit to the factory's habit. We also grade differently by garment: a compression layer grades less than a loose training tee; a structured woven jacket grades more in length than in width. The grade rule has to match the garment's behaviour, not a generic table.
4. Grade the Tolerance
Do not write one flat ±1 cm across all points. We recommend: structural points — chest, body length, sleeve — held at ±1.0 cm; stretch/elastic points — waistband, hem, neck opening — relaxed to ±1.5 cm. Graded tolerance both holds the silhouette and avoids sentencing reasonable elastic recovery to "defect." A flat ±1 cm on 12 points can, in the worst case, let two neighbouring sizes differ by 2 cm on a point and manufacture false rejects. Tolerance grading is how you protect both the fit and the yield.
5. Gate with a Fit Sample (Full Size Set)
Before bulk, produce at least S / M / L / XL — one of each — as fit samples, not just a single signed base size. Worn on a body or a form, the points that drift in the size run surface. This is the only pre-bulk gate that catches fit; skip it and the drift waits for the complaint to expose it. Within our sampling window of 7–10 days (with a complete tech pack), the full size set is included — so the gate does not cost you calendar time, only the discipline to use it. A base-size-only sign-off is a bet that every other size graded itself correctly. It rarely does.
6. Align the Market System
State plainly which system the chart follows — US, EU, UK, or Asian — and attach a cross-system reference. If you sell multi-market, we recommend one system as the master with the others as a conversion appendix, so the chart on the shelf matches what the end consumer expects. This has to be locked before the sealed sample; changing the system at bulk stage is expensive and risks a whole re-grade. When the system is named and the appendix is attached, "off-size" feedback at retail drops because the consumer's expectation and your spec finally speak the same language.
How We Receive Your Size Pack
Saying "send us your size chart" means nothing without the steps. Here is how we turn a partial pack into a grader-ready spec — and it doubles as the list to walk through when you audit any factory's sizing offer:
A Real Size Chart + Grade Rule Template
Plenty of factories say "we can grade." But a grade rule the buyer cannot read is no rule. Below is the format we actually use when we return a grading proposal — the figures below are a format example — illustrative, not a commitment. Use it as a ruler and ask any factory: does your spec carry these fields?
| Field | What it means | Example value |
|---|---|---|
| Base size | Master size all others grade from | M (men's) |
| POM set | Named points of measure | Chest / waist / body length / sleeve |
| Grade rule (woven) | Increment per size step | Chest +4 / length +2 / sleeve +1.5 cm |
| Grade rule (stretch) | Reduced increment for 4-way stretch | Chest +2 cm per size (start) |
| Tolerance (structural) | Allowed deviation, key points | ±1.0 cm |
| Tolerance (elastic) | Allowed deviation, stretch points | ±1.5 cm |
| Market system | Which sizing standard the chart follows | US, with EU/UK appendix |
| Size set for fit | Sizes produced as fit samples | S / M / L / XL |
Note: the figures above demonstrate the format — which fields a grading spec needs and how a woven vs stretch rule differs. Real increments move with the garment type, the fabric's recovery, and your target market, always confirmed against a real fit sample. We do not treat example figures as commitments.
What a Complete Size Chart Actually Looks Like
To make the template concrete, here is a sample grading table for a men's woven training tee, base size M. Read it as a worked example of how a base measurement, per-size grade, and tolerance sit together on one sheet:
| Point of measure | Base M (cm) | L | XL | Tolerance |
|---|---|---|---|---|
| Chest (1 cm below armhole) | 104 | 108 | 112 | ±1.0 cm |
| Body length (CB neck to hem) | 70 | 72 | 74 | ±1.0 cm |
| Sleeve length (shoulder to cuff) | 22 | 23.5 | 25 | ±1.0 cm |
| Waist (relaxed) | 100 | 104 | 108 | ±1.5 cm |
| Hem width | 102 | 106 | 110 | ±1.5 cm |
The same tee in a 4-way stretch fabric would not use +4 / +2 / +1.5. The chest would grade closer to +2 cm per size, because the fabric recovers and a larger chest increment would over-constrict the bigger sizes. That single decision — woven logic vs stretch logic — is the difference between an XL that fits and an XL that gets returned. It is also why our tech pack guide treats fabric behaviour as a first-class spec input, not a footnote.
The Buyer Self-Fill Size & Grading Checklist
Before you send a size pack to any factory, walk this list. Ten minutes of self-audit beats a season of returns — and it works equally well as a checklist to circulate to your team:
The sealed-sample line matters more than it looks. A bulk run can only be judged against an approved reference — the same principle behind our tech pack discipline: the spec you write before production is the spec you enforce after it. Fit and quality are one loop, not two.
With a Grade Rule vs Without — What Changes
| Dimension | Size chart + grade rule complete | Base size only, no grade rule |
|---|---|---|
| Size run consistency | Held to the grade rule, stable across the lot | Every size to factory default, drifts |
| Off-size complaints | Low | Frequently high |
| Cross-size rework | Rare | Large/small sizes re-cut repeatedly |
| Communication rounds | 2–3 | 6–8 and up |
| Hidden cost | Low | Returns, rework, trust all lost |
The pattern repeats across every category we run. A complete spec is not bureaucracy — it is the cheapest insurance on the order. It also shortens the timeline, because the questions that would otherwise surface in production are answered on paper first. Our sportswear lead time guide shows where a clean spec removes days from the schedule.
Questions Buyers Actually Ask
Our size chart only has a base size — can you grade it for us?
Yes. That is exactly the gap we close as an OEM/ODM partner. You send the base size plus the category (woven / knit / stretch), and within 48 hours we return a grade rule proposal — per-size, per-point increments plus graded tolerance. You confirm, and we grade to it. We do not recommend the "you give us one M and we guess the rest" approach, because guessing is where off-size starts.
How do I size stretch and knit fabrics without drifting?
The core move is to halve the grade increment. For 4-way stretch we recommend starting at +2 cm chest per size, where woven men's commonly uses +4 cm. Stretch points also need a looser tolerance, around ±1.5 cm, or the fabric's natural recovery gets misjudged as a defect. We hold 200+ optional fabrics in Dongguan and can measure the actual recovery of the fabric you pick before we fix the grade.
What tolerance should I generally allow?
Structural points — chest, body length, sleeve — at ±1.0 cm; elastic points — waistband, hem, neck opening — at ±1.5 cm. Do not write a flat ±1 cm across all 12 points: in the worst case neighbouring sizes could differ by 2 cm on a point, which manufactures false rejects. Graded tolerance protects both the silhouette and the yield.
Is a fit sample required, and how long does it take?
Strongly recommended, and at least S/M/L/XL, one each. With a complete tech pack, the full fit sample set is included in our 7–10 day sampling window. Skipping it and opening bulk directly leaves the size run's drift risk to the bulk and the complaint — which costs ten times more than the sample.
How do I handle US / EU / UK size differences?
The tech pack must name which system the chart follows and attach a cross-system reference. Among the 1,200+ brands we have worked with, those selling multi-market usually pick one system as master and treat the rest as a conversion appendix, so end consumers do not feel off-size. Lock this before the sealed sample — changing the system at bulk stage is costly.
Does the number of sizes (size run) affect MOQ and unit price?
Yes. MOQ 300 is the total-piece threshold, but a wider size run (say 5 sizes vs 3) thins the pieces per size, so the amortisation of machine setup and marker making rises and the unit price climbs. We suggest new brands start with 3 sizes to test the style, then widen once it sells, rather than laying out 7 sizes from the start.
What does FOB Shenzhen mean — does it include freight?
FOB Shenzhen means "free on board at Shenzhen port" — we carry the goods from Dongguan to Shenzhen port, complete export customs clearance, and load them onto the vessel; all cost and risk up to that point are ours. Ocean freight and insurance from the port to your destination are your side (or your freight forwarder's). The quotation states the split clearly; there is no ambiguity.
Can you work from my existing tech pack, or do I need a new one?
We can work from your existing pack. If it already carries POM, a base size, and a grade rule, we validate it against a fit sample and proceed. If it is thin, our 48-hour grading-gap list tells you exactly what to add before bulk. Either way the sealed sample and graded tolerance are the reference we measure bulk against — the same discipline behind our OEM basics page.
What if the fit sample passes but bulk still drifts?
Then the drift is a production-control issue, not a spec issue — and that is where the inspection half takes over. Bulk is measured against the sealed sample and the graded tolerance under AQL 1.5 with our 8-step process, so a size that leaves tolerance is caught before packing, not after delivery. A good spec plus a good QC loop is what keeps a clean size run clean through the whole order.
Send Us Your Size Pack
Send us your size chart — even if it is only a base size — or a reference image, and within 48 hours we return a grading-gap list, completing the grade rule, POM, graded tolerance, and fit-sample plan. Base size right plus grade rule written: you can hold the first fit sample in 7–10 days.
As a sportswear manufacturer in Dongguan since 2014, we would rather build a long-term partnership than take one order at a time. If anything about size and grading is unclear, ask — we will give it to you straight. To see our OEM base capacity first, visit our OEM basics page.
