Stitch Types and Seam Quality: A Garment Approval Guide

A stitch number identifies a formation; it does not award a quality grade. Garment seam approval needs three separate answers: which stitch and seam construction was specified, whether the formation is continuous and balanced on the approved material, and whether the finished operation meets the project-specific appearance and performance criteria.

That distinction prevents two common errors. A buyer should not approve a seam only because it looks like a 301, 504 or 406. Nor should a buyer reject a location only because its stitches per inch differ from a generic online chart. Fabric, seam type, stitch type, thread, density, tension, finish and end use interact, so the controlling requirement belongs to the specific operation and product.

This article owns stitch identification, formation evidence and the seam-approval record. Thread construction and size, seam puckering, seam slippage, needle damage and bartack placement have their own diagnostic owners.

What ISO 4915 Does—and Does Not—Tell You

ISO 4915:1991 classifies, designates, describes and illustrates stitch types. As of August 12, 2026, ISO lists the edition as published and under systematic review. The number is therefore useful controlled vocabulary, but ISO 4915 is not a product performance standard and does not prescribe a universal garment location, density, thread or pass/fail limit.

The following names are practical starting points for reading a sample. Confirm the exact thread-path diagram and seam assembly in the approved construction record rather than relying on the face view alone.

Stitch designation	Recognizable formation clue	What still needs to be specified
301 lockstitch	One needle-thread line on the face with interlocking needle and bobbin threads	Seam type, location, density, thread, balance, backtack/end security and performance criterion
401 chainstitch	One needle-thread line with a chain formation on the reverse	Seam geometry, extension need, end security, density and thread combination
504 overedge stitch	Thread formation encloses a cut edge	Whether it only finishes an edge or participates in a seam, bite/width, density and thread setup
516 safety-stitch combination	A chainstitch seam and overedge formation appear in one construction	Component stack, seam allowance, spacing between formations and operation-specific criteria
406 covering stitch	Two needle-thread rows on one side with a covering looper formation on the other	Hem geometry, fold, needle spacing, density, balance, extension and appearance
607 flatseam-family formation	Multiple threads cover a flat or flattened join	Exact edge/seam preparation, coverage, bulk, appearance and performance requirement

Do not identify “flatlock” from a sales label alone. The term can be used loosely for visually flat constructions. Photograph both faces and the cross-section or seam allowance, then name the ISO stitch and seam assembly that the project actually approved.

Stitch Type, Seam Type and Operation Are Different Fields

A stitch is the repeated thread formation. A seam is the way material plies and edges are arranged and joined. An operation adds location, material stack, machine setup, handling and finishing exposure. The same stitch type can appear in different seam assemblies, and similar-looking seams can use different formations.

ASTM D6193-16(2025) is a standard practice for stitches and seams. Its public summary states that seam performance depends on variables including fabric type, seam and stitch type, stitch density, thread tension and strength, and product end use. This is why “use overlock” or “use 12 SPI” is incomplete purchasing language.

Write the approval unit as a location-specific operation, for example:

Operation ID and garment location:
Material identity, face and ply stack:
Seam type / edge preparation:
Stitch type and machine configuration:
Needle and looper thread identities:
Needle system/size where controlled:
Stitch density and how/where measured:
Seam width, bite, allowance or needle spacing:
End/junction security:
Face and reverse appearance criteria:
Required extension or load direction:
Finish/wash route before final assessment:
Named test method, specimen direction and acceptance criterion:
Approved sample ID and revision:
Disposition authority:

That block is more reproducible than copying one stitch number into a tech pack without saying where or how it is used.

There Is No Universal SPI Quality Grade

Stitches per inch, or SPI, is one process and construction variable. It can affect the amount of thread, the number of penetrations, the way load is shared and the visual scale of the operation. It also interacts with stitch length, material structure, needle/thread relationship, tension, seam geometry and sewing speed.

For that reason, a higher count is not automatically better and a lower count is not automatic evidence of cost cutting. Excessive density can add needle penetrations or material damage; insufficient density may not deliver the approved formation or performance. The correct value is the one validated for the actual operation and recorded with a tolerance and measurement method.

When measuring density:

1. name the operation and specimen state;
2. choose a straight, representative section away from starts, ends and junctions unless those are the subject of the check;
3. state the measurement length and how partial stitches are handled;
4. record more than one location when the project sampling plan requires it; and
5. compare with the approved specification, not an uncited garment-category chart.

If the factory changes thread, needle, material, finish or seam geometry, do not assume the old density proves equivalent performance. Make a controlled trial and update the approved record where the change is accepted.

Separate Skipped, Broken and Imbalanced Stitch Evidence

These conditions can all interrupt the expected visual rhythm, but they are not interchangeable.

Evidence category	What to look for before assigning cause	What the evidence does not prove
Skipped stitch	An expected loop/interlock was not formed; the line may show an abnormally long interval or a thread that did not connect with its partner	It does not by itself identify timing, clearance, needle, thread control, flagging or another setting as the cause
Broken stitch or thread	A previously formed line is interrupted by visible broken ends, loss of continuity, opening or runback	It does not establish whether breakage occurred during sewing, finishing, handling, inspection or use
Imbalanced formation	Interlocking or looping sits visibly to one side, or needle/looper threads are not seated in the approved relationship	It does not prove that changing only the top tension will solve the condition
Open seam	Material edges separate because the stitch formation, thread or seam system no longer holds the intended join	It does not automatically distinguish thread rupture from seam slippage or fabric failure

American & Efird's Common Seam Quality Defects bulletin describes a skipped stitch as a stitch-forming device failing to pick up a loop, and distinguishes broken or restitched seams and open-seam stitch failure. Coats' sewing-problem guidance lists multiple possible contributors to skipped and imbalanced stitches, including machine timing/clearances, needle-thread relationship, thread control, material flagging or deflection, threading and tension. Groz-Beckert's skipped-stitch technical note likewise explains the missing loop/hook connection and shows why machine setting and needle condition must be investigated rather than inferred from one photograph.

Use those sources as diagnostic maps, not as permission to change several variables at once.

Build an Evidence Chain Before Correcting the Seam

Start with non-destructive observations on the identified sample:

1. Bound the location. Record garment/sample ID, operation, side, start/end or junction and the condition before handling it.
2. Photograph both faces. Include scale, stitch direction and nearby construction landmarks.
3. Map continuity. Mark the first and last visibly affected stitch and whether the thread path resumes.
4. Inspect thread ends and material. Record broken ends, missing loops, displaced fabric yarns, holes, abrasion or an opened seam without assuming sequence.
5. Compare with a known conforming operation. Use the same material, ply stack, direction and finish state where possible.
6. Review machine and material records. Check threading, needle, thread packages, settings, maintenance or change events and operator notes.
7. Run one controlled correction. Change the variable named by the technical owner, keep other relevant inputs fixed and sew an identified trial.
8. Reassess after the intended finish. A formation that looks acceptable before wash, dye, abrasion or extension may not represent the finished garment.

Destructive opening, extension or laboratory testing should use a designated specimen and named method. Do not pull an approval sample until it fails and call that an objective seam test. Preserve the original evidence and state whether the assessed piece was production-intent, retained, rejected or made solely for diagnosis.

AMANN's needle-thread tension guidance describes evaluating thread balance and warns that settings are transferable only under identical machine conditions; thread and seam factors still matter. That supports a controlled comparison, not a universal tension number.

Match the Test Method to the Failure Question

Visual approval and mechanical testing answer different questions. Name the material scope, specimen preparation, seam direction and loading mode before citing a result.

Reference	Publicly stated scope	What it cannot prove alone
ISO 4915:1991	Classification and terminology for stitch types	Seam strength, durability, appearance or product acceptance
ASTM D6193-16(2025)	Standard practice for stitches and seams	One universal stitch, density or tolerance for every garment
ASTM D1683/D1683M-22	Failure in sewn seams of woven fabrics under force perpendicular to the seam	Actual wear life or suitability outside its stated material/specimen scope
ISO 13935-1:2014	Strip method for seam maximum force, mainly for woven textile fabrics and straight seams	Multidirectional use behavior or every garment seam geometry
ISO 13935-2:2026	Grab method for seam maximum force, mainly for woven textile fabrics and straight seams	A universal finished-garment acceptance limit

Do not substitute a fabric burst result for a sewn-seam method merely because the garment is knitted. Do not turn a test method into a pass/fail number unless the project's requirement, material scope and specimen condition define that number.

If the visible issue is fabric yarn displacement next to an intact woven seam, route the evidence to the seam-slippage owner. If the seam line is wavy or gathered while the formation remains continuous, use the seam-puckering diagnosis. If holes, cut yarns or runs appear near needle penetrations, use the needle-damage owner.

A Buyer-Side Seam Approval Record

For each critical or appearance-sensitive operation, the approval packet should connect:

• the tech-pack operation and construction drawing;
• actual material, layer stack and direction;
• stitch and seam designations;
• thread identities and any controlled needle information;
• measured density and other geometry using the stated method;
• face, reverse, junction and end-security photographs;
• finish or wash state;
• skipped, broken, imbalanced, pucker, damage or slippage observations;
• required mechanical method and result where applicable;
• approved sample/revision; and
• pass, revise, reject or hold decision with bounded scope and authority.

One close-up is not the full record. A macro photograph can show formation but hide garment location; a full-garment image can show location but hide individual loops. Keep both and use stable IDs.

The same evidence can then support later inspection without pretending that visual workmanship sampling replaces laboratory performance testing. The garment AQL guide owns sampling-plan language, while the sample approval checklist owns the disposition of the development sample as a whole.

FAQ

Does an ISO stitch number mean the seam is high quality? No. ISO 4915 supplies classification and terminology. Quality depends on the specified seam assembly, material, thread, density, setup, finish, appearance and performance criteria for that operation.

What SPI should a brand specify for a T-shirt or hoodie? There is no universal number for the entire garment category. Validate density by operation on the actual material and construction, then record the target, tolerance and measurement method. A neck seam, hem and decorative topstitch need not share one value.

How can I tell a skipped stitch from a broken stitch? A skip is evidence that an expected loop or interlock did not form; a break interrupts thread that may already have formed stitches. Photograph both faces, preserve the ends and trace continuity before deciding when or why the failure occurred.

Can I approve seam strength by pulling the sample by hand? A gentle handling check may reveal an obvious open seam, but it is not a reproducible strength test. Use a designated specimen, named method, direction and project acceptance criterion when mechanical performance is required.

Is a balanced-looking seam automatically approved? No. Balance is one formation observation. The seam must also match the approved operation and meet the applicable appearance, extension, end-security, finish and performance requirements without introducing damage or another defect.

Primary Sources

• ISO, ISO 4915:1991 — stitch-type classification, designation, description and illustration; status checked August 12, 2026.
• ASTM International, ASTM D6193-16(2025) — stitches/seams practice and interacting performance variables.
• American & Efird, Common Seam Quality Defects — formation evidence for skipped, broken/restitched and open seams.
• Coats Group, Solving Your Sewing Problems — multi-variable diagnostic guidance for skipped and imbalanced stitches.
• Groz-Beckert, Skipped Stitches technical note — loop pickup, needle and machine-setting evidence.
• AMANN, Needle-thread tension for balanced stitches — balance assessment and limits on transferring settings.
• ASTM International, ASTM D1683/D1683M-22 — woven sewn-seam failure/strength method and stated limits.
• ISO, ISO 13935-1:2014 and ISO 13935-2:2026 — strip and grab methods for seam maximum force.