A specification such as 45 × 45 / 96 × 72 contains two separate measurements: yarn count and fabric density. They describe different parts of the construction. If a buyer or supplier reads the figures in the wrong order, a fabric can be approved against the wrong requirement even if it looks similar at first glance.
This guide explains ends per inch (EPI), picks per inch (PPI), how to check them, which recognised test methods apply, and how buyers should record results. It is written for woven fabrics; knitted fabrics use different construction terms.
1. What do EPI and PPI mean?
- Ends per inch (EPI): the number of warp yarns across one inch of fabric width.
- Picks per inch (PPI): the number of weft yarns along one inch of fabric length.
The warp runs in the length direction of the fabric roll. The weft, also called filling, runs across the width. One warp yarn is an end; one inserted weft yarn is a pick. The two values must be reported separately because a fabric can have a different density in each direction.
In metric reporting, results may be expressed as ends per centimetre and picks per centimetre. Always state the unit. A value of 38 ends/cm is not the same as 38 EPI.
2. Read the fabric construction specification correctly
A common woven-fabric notation is 45 × 45 / 96 × 72. In a typical cotton-count specification, the first pair refers to warp and weft yarn count, while the second pair refers to warp ends and weft picks per inch. The yarn-count system must still be confirmed: do not assume that “45” means Ne unless the buyer’s specification says so.
| Specification element | What it describes | What to confirm |
|---|---|---|
| 45 × 45 | Warp and weft yarn count | Count system, such as Ne, and whether the yarns are single or plied |
| 96 × 72 | Warp ends × weft picks | Whether values are per inch or per centimetre |
| GSM | Fabric mass per unit area | Test method and whether the value is for greige or finished fabric |
| Width | Fabric width | Overall, usable or cuttable width, and the measurement stage |
EPI and PPI are not yarn count. They are not GSM, and they should not be combined into one number. For example, adding 96 + 72 and reporting “168” loses the separate warp and weft information needed to verify the ordered construction.
For related explanations, see our guides to yarn-count systems and fabric GSM calculation.
3. Which standards cover woven-fabric thread count?
Use the test method named in the buyer’s specification or agreed quality plan. Two recognised references are:
| Reference | Scope and method information | Practical point for buyers |
|---|---|---|
| ISO 7211-2:2024 | Specifies three methods for determining threads per centimetre in woven fabrics: Method A, fabric dissection; Method B, counting glass; and Method C, a traversing thread counter. Method A takes precedence in case of dispute. | Method A is suitable for all fabrics; Method B is suitable for fabrics with more than 50 threads/cm; Method C is suitable for all fabrics. |
| ASTM D3775-17(2023) | Standard Test Method for End (Warp) and Pick (Filling) Count of Woven Fabrics. It covers end and pick count for woven fabrics. | The ASTM page identifies it as suitable for acceptance testing of commercial shipments. Confirm the edition and units required by the contract. |
Important edition note: ISO 7211-2:2024 is the current published edition and replaced ISO 7211-2:1984, which was withdrawn. Avoid writing the old edition into a new test request unless a buyer’s controlled specification explicitly requires it.
A pick glass or counting glass is useful for quick visual checks, but the tool alone does not establish that a formal test has followed every requirement of a named standard. In particular, ISO 7211-2:2024 describes Method B as suitable for fabrics with more than 50 threads per centimetre. A fabric specified at 96 × 72 per inch is approximately 37.8 ends/cm × 28.3 picks/cm, so do not automatically assume that Method B is the appropriate ISO method for that construction. Select the method based on the fabric and the standard’s scope.
4. How to check EPI and PPI on a fabric sample
Step 1 — Confirm the order specification
Before counting, confirm the warp and weft directions, target construction, unit, yarn-count system, fabric stage (greige, dyed or finished), test method and buyer-agreed tolerance. If the specification only says “96 × 72” without stating the unit, clarify it before making an acceptance decision.
Step 2 — Identify warp and weft
Use the roll direction, selvedge and fabric construction to identify the warp and weft. Mark the directions on the sample. Do not assume that a visible stripe always follows the warp; check the actual yarn direction.
Step 3 — Choose a suitable counting method
For a quick in-process check, a suitable pick glass can help an inspector see individual yarns. For a formal report, use the method required by the purchase specification or applicable standard. Depending on the fabric, this may involve dissection and direct counting or a suitable thread-counting instrument.
Step 4 — Take representative readings
Lay the specimen flat and free from folds or tension. Do not stretch it to make the yarns easier to see.
Avoid counting only at the selvedge or selecting one unusually neat area. Take readings at representative positions across the usable width and along the roll, following the agreed sampling plan. ISO 7211-2:2024 calls for counts at at least five different places and specifies a minimum measuring distance according to thread density:
| Thread density | Minimum measuring distance |
|---|---|
| Below 10 threads/cm | 10 cm |
| 10 to 25 threads/cm | 5 cm |
| 25 to 40 threads/cm | 3 cm |
| Above 40 threads/cm | 2 cm |
For Method A (fabric dissection), the standard requires a specimen containing at least 100 threads. These are method requirements, not a substitute for the buyer’s lot-sampling plan. Keep individual readings so local variation is not hidden by an average.
Step 5 — Count each direction separately
Count warp ends across the measured width for EPI, then count weft picks along the measured length for PPI. Follow the selected method’s rules for the measurement boundaries and minimum measuring distance. On a patterned or striped fabric, count yarns—not colour bands, printed lines or design repeats.
Step 6 — Record the result and compare it with the agreed limit
Record each result, the measurement unit, test method, location, fabric stage, instrument and any conditioning or preparation required by the method. Compare the measured construction with the purchase specification and its stated tolerance. Do not invent a universal tolerance: acceptance limits must come from the buyer’s approved specification or contract.
5. Worked example: convert a measured length to EPI and PPI
Suppose an illustrative check counts 189 warp ends across a 5 cm measuring length and 142 weft picks along a 5 cm measuring length. To convert a count per centimetre to a count per inch, multiply by 2.54 because one inch equals 2.54 centimetres.
EPI = (189 ends ÷ 5 cm) × 2.54 = 96.012 ends/in, or approximately 96.0 EPI.
PPI = (142 picks ÷ 5 cm) × 2.54 = 72.136 picks/in, or approximately 72.1 PPI.
The illustrative result is therefore approximately 96.0 × 72.1 per inch. For a 96 × 72 per-inch construction, the converted densities are about 37.8 ends/cm and 28.3 picks/cm; both fall in the 25–40 threads/cm band, for which ISO 7211-2:2024 lists a 3 cm minimum measuring distance. The 5 cm example is longer than that minimum, but the method and buyer sampling plan still control; the calculation is not an acceptance rule. Use the measuring distance and counting procedure specified by the chosen standard, and make the acceptance decision against the buyer’s agreed tolerance.
6. What can a density difference indicate?
A lower-than-specified EPI or PPI may indicate a construction setting or production-control issue, a sample or measurement problem, or a difference between the greige and finished measurement stages. Possible investigation points include the loom construction settings, warp preparation, weft insertion and take-up records, yarn breaks, fabric relaxation and finishing history. These are investigation leads, not proof of a particular cause.
Thread density can influence fabric cover, appearance, opacity, handle, mass and performance. However, a higher EPI or PPI does not automatically mean better fabric. Yarn linear density, weave structure, yarn diameter, crimp, fibre type and finishing all affect the final result. Review construction together with GSM, width, appearance and the performance tests relevant to the end use.
If the fabric is measured at different production stages, identify those stages clearly. A greige result and a finished result should not be compared as if they were measured on identical material under identical conditions. Our greige-fabric inspection checklist explains the checks to complete before dyeing and finishing.
7. Common EPI and PPI measurement mistakes
- Mixing units: comparing ends/cm with EPI or picks/cm with PPI without converting.
- Confusing yarn count with density: yarn count describes yarn fineness; EPI and PPI describe how many yarns occupy a measured length.
- Counting only one location: one reading may miss variation across the width or along the roll.
- Counting a pattern instead of yarns: colour stripes, printed lines and weave repeats are not substitutes for counting individual ends and picks.
- Using an unsuitable method: a counting glass may not be suitable for every construction or density; follow the chosen standard.
- Reporting only an average: retain the individual readings when local variation matters.
- Assuming a universal tolerance: apply the approved buyer requirement, not an informal industry-wide percentage.
8. What buyers should include in an RFQ or inspection report
To avoid disputes, state the construction requirement in a complete and measurable way. A practical RFQ or inspection record should include:
- Warp yarn count and weft yarn count, including the count system.
- Target EPI and PPI, with units stated explicitly.
- Applicable test method and edition, such as ISO 7211-2:2024 or ASTM D3775-17(2023).
- Agreed sampling plan, measuring locations and acceptance tolerance.
- Fabric stage: greige, dyed, finished or otherwise processed.
- Individual warp and weft readings, calculated results and any variation by location.
- Roll or lot identification, inspection date, instrument and inspector/laboratory details.
- Related requirements such as GSM, usable width, composition, weave, shade and end-use performance.
If a result is outside the agreed limit, first confirm the sample identity, method, unit and calculation. Then arrange a repeat or expanded check under the agreed procedure and investigate the relevant production records. Keep the original readings and the follow-up result together so the decision is traceable.
Conclusion
EPI and PPI are simple terms, but a reliable construction check depends on correct direction identification, the right unit, a suitable method and representative measurements. State warp ends and weft picks separately, use the standard edition agreed with the buyer, and never treat an average or an assumed tolerance as a substitute for the purchase specification.
For a fabric quotation or technical review, share the construction, yarn-count system, composition, GSM, width, quantity and intended application through our contact page.
Standards and image credit
- ISO 7211-2:2024 — Textiles: Methods for analysis of woven fabrics construction, Part 2.
- ASTM D3775-17(2023) — End (Warp) and Pick (Filling) Count of Woven Fabrics.
Featured image credit: “Blue chambray cloth,” photograph by Edward G. FitzGerald, 29 January 2012, via Wikimedia Commons, licensed under CC BY-SA 3.0. The image is an illustrative photograph of chambray fabric with a millimetre scale; it is not a pick-glass reading or a measured result for the worked example.



