Journal/Starting out

How to Build a Size Chart for Your Clothing Brand

A tape measure beside three folded T-shirts in different sizes and a notebook with a blank grid chart.

Sizing is where brands quietly lose money. Bad fit drives returns, kills reviews, and erodes trust faster than almost anything else. A clear, well-built size chart is how you prevent it — and it's more straightforward than it looks.

The building blocks

  • Points of measure (POM) — the exact places you measure a garment: chest, body length, sleeve, shoulder, hem, and so on.
  • Base size — the size you design and sample first (often M). Everything else grades off it.
  • Grading — the increments you add or subtract between sizes at each POM.
  • Tolerances — the allowed variance in production (for example ±1 cm), because no two garments are ever identical.

How to build one, step by step

  • Pick your base size and your size range (e.g. XS–XXL).
  • List your points of measure for the garment type.
  • Measure a fit sample — or a reference garment whose fit you love.
  • Apply your grading increments across the range at each POM.
  • Add tolerances so production has a realistic target.
  • Wash-test — fabric shrinks, so measure after washing, not just off the roll.

Standard vs custom sizing

You can start from a standard body-measurement chart (ISO and ASTM both publish standard body sizes) and adjust to your fit and audience — a streetwear brand runs oversized, a fitted women's line runs closer to the body. Watch regional differences too: US, UK and EU size numbering don't match, so state which you're using.

Common mistakes

  • Designing without a defined base size, so nothing grades cleanly.
  • Forgetting shrinkage — always wash-test before finalising.
  • Tolerances too tight to manufacture, or so loose that sizes feel inconsistent.
  • Copying a competitor's chart without checking it against your own fit sample and fabric.

We give you a base size chart to work from and grade it to your fit, and every style is sampled and wash-tested before bulk. Sizing matters most where bodies change fastest, which is why kidswear gets its own graded chart. See how sampling works and what goes in a tech pack.

Frequently asked questions

How do I create a size chart for my clothing brand?

Work in sequence, because each step depends on the one before. Choose a base size — usually your expected best-seller, often M — and the range you will offer. List your points of measure: chest, body length, shoulder, sleeve, hem, and any others the garment needs. Measure your approved fit sample at each point, which gives you the base column. Apply grading increments to generate the sizes above and below. Add tolerances per measurement, since no factory produces identical garments. Then wash-test, because a chart built on unwashed measurements describes a garment your customer will never own. Starting from a published body chart such as ISO 8559 and adjusting toward your own fit is faster than deriving one from nothing.

What is grading in clothing?

Grading is the rule set that generates every other size from your base size — the specific amount added or removed at each point of measure between one size and the next. It is not uniform scaling, which is the most common misunderstanding: a chest measurement might grow by a larger increment than a shoulder, because bodies do not enlarge proportionally in every dimension. Sleeve and body length usually grade in smaller steps than width. Consistent grading is what makes a range feel coherent, so a customer who takes an L in one of your styles takes an L in all of them. Inconsistent grading produces the experience of a brand whose sizing cannot be trusted, which drives returns more effectively than a bad fit in a single style.

What are points of measure?

Points of measure are the specific, agreed locations where a garment is measured, and the value of defining them is that they remove interpretation. "Chest" alone is ambiguous; "chest measured one inch below the armhole, laid flat, edge to edge" is a specification two people in different countries will apply identically. A typical knit top uses chest, body length from high point shoulder, shoulder width, sleeve length and hem width, with more added for complex garments. Your size chart lists a target for each point at every size. When a garment is later inspected, it is these measurements that are checked — so a vague point of measure becomes a quality dispute nobody can resolve objectively.

What tolerance should a size chart have?

A defined variance per measurement, commonly in the region of one centimetre, though it should vary by point rather than being applied uniformly. Fabric moves, particularly knits, and cutting and sewing introduce small differences no factory can eliminate — so tolerance is an acknowledgement of physics rather than a concession. Set it too tight and you create an unmanufacturable specification that generates rejections over differences no customer would notice. Set it too loose and sizes stop feeling consistent, which is worse for returns than a slightly unusual fit. Larger measurements generally warrant more tolerance than small ones. Ask your factory what is realistic for your specific fabric, then write the agreed figures into the tech pack so inspection has an objective standard.

Free tools & guides for this

Sources & further reading

Have an idea? Let’s make it.

We manufacture from 100 pieces per style, with GOTS-certified organic options and photos at every stage. Send a sketch or a sentence — we’ll reply within a day.

Start your collectionMore guides