Used in apparel, home textiles & garment accessories.
Dedicated to in-depth expertise and continuous innovation in knitted textiles.
A brand sends a tech pack for a seven-eighth training tight, receives the first sample, and the fabric photographs well but feels slick and papery in the hand. The second round fixes the hand feel, then goes sheer across the knee the moment the leg is bent. The third round passes the fitting, ships to bulk, and comes back from the wash test half a size short with a dull, fuzzy surface. Three months gone, and nothing was wrong with the design.
The pattern was never the problem. The fabric was chosen by description instead of by number. Athletic apparel fabric behaves as a system: the knit structure decides how it moves and where it holds shape, the fibre blend decides how it feels and how fast it dries, and the finishing decides whether it still looks like the sample after twenty washes. Line those three up early and most of the painful sample rounds simply do not happen.
This guide is written from the mill side of that conversation. Shaoxing Keqiao Maozan Textile Co., Ltd. is a knitted fabric manufacturer based in Keqiao, the Shaoxing district next to the China Textile City cluster, working across single jersey, rib, interlock, air layer, ponte roma, pique, mesh, waffle, terry, fleece and crepe for apparel, home textiles and garment accessories. What follows is how those structures are usually matched to activewear programmes, and which numbers are worth fixing in a purchase order before the first metre is knitted.
Athletic apparel fabric is judged on stretch recovery, opacity under tension and moisture behaviour long before anyone judges the colour. Fix those three in writing and the colour conversation becomes the easy one.
Content
Activewear is unusual in that the customer tests the fabric with their body within ten seconds of putting it on. A t-shirt can be slightly stiff and nobody notices. A legging that pulls at the waistband, shines across the thigh or turns transparent in a squat is returned, whatever the print looks like. So the specification conversation has to start with function.
There are four jobs almost every performance knit has to handle at the same time:
Hand feel sits on top of these four, not underneath them. A brushed, peach-skin surface on a training top feels wonderful in a showroom and pills badly after fifteen washes unless the fibre choice and the raising process were designed together. When a buyer asks for "soft but technical", the honest answer is usually a specific blend, a specific gauge and a specific finishing route, not a single magic yarn.
Performance is a combination of structure, blend and finishing; any one of the three chosen in isolation will show up as a complaint at the fitting stage.
Weft knitting gives an enormous range of behaviour from the same yarn, purely by changing how the loops interlock. For activewear, six structures do most of the work, and a seventh group covers warm-up and lifestyle pieces.
One set of loops, one face, one back. Single jersey is thin, drapes well, breathes easily and is cheap to knit, which is why it dominates training tees, tank tops and light liners. Its weakness is stability: it curls at the edges, can spiral after washing, and stretches more in the course than in the wale. A cotton/spandex jersey at 160–180 GSM is the classic gym tee; a 100% polyester jersey at 130–150 GSM with a wicking finish is the classic running tee. Anything below about 120 GSM starts to feel insubstantial and becomes hard to cut and sew without distorting.
Two sets of loops, both faces smooth. Interlock is denser, heavier and far more dimensionally stable than jersey, which makes it the default for leggings, sports bras, cycling shorts and fitted tops. The important detail is that interlock holds its shape without needing heavy spandex loading, so a 220–260 GSM nylon/spandex interlock can deliver firm support at 18–22% elastane while still feeling smooth on the inside.
Nylon-Spandex Interlock Fabric for Compression ActivewearDense, smooth double-knit with high stretch and recovery for leggings, sports bras and fitted tops; composition, weight and OEKO-TEX options are configurable.View Product →
Nylon brings a cool, slightly slick hand and excellent abrasion resistance, which matters on the inner thigh of a tight. Polyester interlock costs less and dries faster but feels drier and less luxurious against skin. Both are legitimate; the choice is usually driven by the retail price point and by how much the customer will notice the inside surface.
Rib is knitted with alternating wales of face and reverse loops, so it contracts laterally and opens up comfortably. In activewear it appears as waistbands, cuffs, neckbands, side panels and full fitted tops. A 1x1 cotton/spandex rib at 220 GSM makes a soft, forgiving waistband; a 2x2 or 4x2 modal rib gives a deeper, more decorative column and a smoother hand. The specification that matters here is recovery after repeated extension, because a waistband that grows is a returned garment no matter how good the body fabric is.
Air layer fabric is a double-face knit with a spacer yarn or tuck structure between two surface layers, trapping a thin cushion of air. It reads as substantial without being heavy, holds a clean silhouette and insulates slightly, which is why it works for hoodies, joggers, gilets, yoga tops and transitional outer layers. A polyester/spandex air layer stretch jersey at around 240–300 GSM gives four-way stretch with a firm, dry hand and a matte face that prints and dyes cleanly.
Polyester-Spandex Air-Layer Stretch Jersey for SportswearCushioned air-layer knit with stretch and recovery for sportswear, hoodies and casual layers; weight, color and finishing can be customized.View Product →
Softer versions built on viscose, modal or a polyester/lyocell blend drape more like a sweatshirt and suit lifestyle and studio-to-street ranges. Slub air layer, where a deliberately irregular yarn creates a subtle vertical texture, is one of the few structures that adds visual interest to a solid-colour programme without printing.
Mesh is an open structure with visible holes, used for ventilation panels behind the knee, under the arm, down the spine of a jersey or across the back of a cap. Polyester mesh knit is light, strong and cheap to knit, and it can be engineered for a specific openness by changing the needle arrangement. The practical risk is that mesh panels rarely match the body fabric in shrinkage, so the seam between them can pucker after the first wash unless both are finished on the same route.
Pique, with its raised waffle-like cells, is technically a ventilation structure too, but in activewear it usually appears as a face fabric for polos, golf shirts and heavier training tops where a smooth jersey would look too plain.
These structures handle the parts of an activewear line that are worn before and after the session rather than during it. Waffle knits give texture and loft, and a cotton waffle or CVC waffle with a little stretch makes a convincing thermal layer. Terry, particularly a 100% cotton or viscose terry, is the base for sweatpants and hoodies. Fleece, including a polyester/spandex fleece or a viscose suede-effect fleece, covers the brushed-back category. Crepe, either a bubble crepe or a textured crepe, brings a crinkled surface that hides creasing and travels well, which is why it turns up in lifestyle and post-workout pieces.
The knit structure sets the ceiling on performance; choosing interlock for a compression tight and jersey for a loose tee is not a stylistic preference but a functional requirement.
Elongation is the number most often left vague in a tech pack, and the vagueness surfaces as a fit complaint months later. The ranges below reflect what buyers commonly write into fit specifications for different garment categories, measured with a strip test on a tensile machine rather than by pulling the fabric between two hands.
Typical elongation ranges written into activewear fit specifications
Two things are worth noticing in that chart. First, the gap between a compression tight and a jogger is enormous, which is why a single "four-way stretch" fabric cannot serve a whole line; the leggings need an interlock or a tightly knitted double knit with 18–22% elastane, while the jogger is comfortable in an air layer or a french terry with 3–6% elastane. Second, the ranges overlap, so two suppliers can both claim compliance and still deliver noticeably different garments.
That is why the elongation figure should always travel with a recovery figure. A practical specification is something like "minimum 80% elongation in the course direction, maximum 5% growth after five cycles of 80% extension and relaxation". This is where a knitted fabric supplier with in-house knitting and a stable yarn supply has an advantage: the elastane percentage, the plaited structure and the heat-setting conditions can all be adjusted together, rather than being bought in as a finished roll and accepted as it comes.
It is also worth specifying stretch direction properly. Many knits stretch more in the course than in the wale, and a legging cut with the greatest stretch running vertically will feel supportive at the waist and loose at the knee. Before approving a bulk order, mark the course direction on the sample and confirm it matches the marker.
Elongation without a recovery limit is half a specification; write both numbers, plus the test method, or two compliant rolls will still sew into two different garments.
Once the structure is fixed, the blend decides almost everything the customer will feel. The table below compares the combinations that appear most often in athletic apparel programmes.
| Blend | Hand feel and drape | Moisture behaviour | Recovery and durability | Typical activewear use |
|---|---|---|---|---|
| Polyester / spandex, 85/15 to 90/10 | Dry, light, slightly crisp; can feel synthetic if not finished softly | Fast wicking, very fast drying | Good recovery, excellent abrasion resistance, holds colour well | Running tees, training tops, mesh panels, light leggings |
| Nylon / spandex, 78/22 to 82/18 | Smooth, cool, with a subtle sheen; the most "premium" synthetic hand | Wicks well but dries slightly slower than polyester | Best-in-class abrasion resistance and compression recovery | Compression tights, sports bras, cycling shorts, swim-adjacent pieces |
| Cotton / spandex, 92/8 to 95/5 | Soft, matte, familiar against skin | Absorbs moisture and holds it; slow to dry | Moderate recovery, prone to stretching at the knee | Low-intensity training, athleisure tees, rib bands and cuffs |
| CVC, roughly 60% cotton / 40% polyester | Softer than pure polyester, more body than pure cotton | Balanced; dries faster than cotton alone | Better shape retention than cotton, moderate pilling risk | Waffle thermals, heavier tees, warm-up tops |
| Viscose, modal and lyocell blends | Fluid, cool, with a soft peach surface; drapes rather than holds | Absorbs well, dries moderately | Lower abrasion resistance, so usually blended with polyester or nylon | Yoga tops, lifestyle leggings, air layer hoodies, terry sets |
| TR and NR ponte roma | Firm, dense, with a clean tailored face | Moderate wicking; more about shape than sweat | Excellent dimensional stability, resists bagging | Studio-to-street trousers, skorts, structured jackets |
The most common blend mistake in activewear is over-indexing on cotton because it feels good in the sample room. A 95/5 cotton/spandex jersey is genuinely comfortable for a low-impact top, but put it on a high-sweat training garment and the customer will feel damp for twenty minutes after the session. If a brand wants the cotton hand with better drying, a CVC or a cotton/polyester/viscose blend is a more honest answer than a heavy wicking finish on a pure cotton base.
The second most common mistake is treating elastane as a dial that only adds stretch. Above roughly 20%, elastane starts to affect cost, dyeing behaviour and heat-setting requirements, and the fabric becomes harder to stabilise. Below about 5%, the garment relies on the knit structure for shape rather than on the yarn, which is fine for a ponte roma trouser and not fine for a compression tight.
Match the blend to how much the wearer will sweat, not to how the swatch feels on a desk; nylon for support and abrasion, polyester for drying speed, cotton and viscose for comfort in low-output pieces.
A one-page specification prevents more arguments than a dozen emails. The table below is a workable starting template for a performance knit, and it is deliberately written in the language a knitting mill uses when it sets up a machine.
| Parameter | What to write | How it is checked | Why it matters |
|---|---|---|---|
| Weight | GSM with a tolerance, for example 240 GSM plus or minus 5% | Circular cutter sample, conditioned and weighed | Drives cost, opacity, drape and the whole fit of the garment |
| Width | Open width or tubular, with tolerance | Measured after relaxation, not straight off the machine | Affects marker efficiency and therefore real fabric consumption |
| Elongation | Minimum percentage by direction | Strip test, recorded at a fixed load | Determines whether the garment can be put on comfortably |
| Growth and recovery | Maximum growth after repeated extension | Cyclic extension, measured after relaxation | Controls bagging at knee, elbow and waistband |
| Shrinkage | Maximum percentage after a defined wash cycle | Wash test to an agreed standard | Prevents the half-size-short complaint after the first home wash |
| Colourfastness | Rating for washing, rubbing and light | Standard test methods on a dyed sample | Protects pale garments from dark trims and vice versa |
| Pilling | Minimum rating after abrasion | Random tumble pilling test | Separates a fabric that looks new for a season from one that looks new for a year |
| Spirality | Maximum twist after washing | Measured on a washed square | Critical for single jersey, where side seams can rotate visibly |
| Roll length | Fixed length per roll with a stated splice policy | Measured at inspection | Protects the cutting room from unexpected short rolls and off-cuts |
Two of these deserve extra attention. Spirality is the quiet killer of single jersey programmes: a tee can pass every physical test and still arrive with side seams that drift forward after washing, because the loops were twisted during knitting and the finishing did not fully relax them. The fix is in the knitting and the relaxation step, so it has to be raised with the knitter, not with the cutter.
Opacity is the other one. There is no universal test, so the practical approach is to agree a fabric, a colour and a stretch condition, then compare against a retained physical standard under the same lighting. For a deeper look at how knit structures, GSM and stretch interact, there is a useful reference on knitted fabric structures, GSM and stretch that covers the machine-level detail behind these numbers.
A signed one-page specification with tolerances, test methods and a retained physical standard removes most bulk disputes before they start.
Most failures in athletic apparel fabric are not dramatic. They are small deviations that only become visible when the garment is worn, washed or photographed under studio lighting. These are the ones that come up repeatedly.
Dark navy, charcoal and deep olive are the hardest to repeat. A second bulk lot that is half a shade lighter will read as a different product on a shelf. Ask how the mill handles lot continuity, whether the greige comes from the same source, and whether a lab dip can be retained as the standard for the whole season rather than per order.
Knitted fabric relaxes. If the finishing route does not allow for relaxation before cutting, the garment will move after the customer washes it. The safe practice is to wash-test the bulk fabric, not just the sample, because a sample from a small trial run and a bulk roll from a full machine load are not the same material.
A fabric can pass a strip test and still fail at the seam, because the stitch type and the seam allowance consume part of the available stretch. Test the sewn seam, not only the flat fabric, particularly on leggings where the inside leg seam takes continuous load.
Anything with a raised face carries pilling risk. The usual causes are short-staple yarns, excessive raising and blends that mix fibres of very different tenacity. If the design calls for a peach-skin hand, agree the pilling rating at the sampling stage, because it cannot be fixed later.
A 5% GSM tolerance sounds tight until it is applied across four colours and three deliveries. One shade may come in at the low end and another at the high end, which changes drape and opacity between colourways of the same style. Buyers who ask for a tighter window on the base shade usually save themselves a fitting problem.
Test the bulk roll and the sewn seam, not the sample swatch; almost every sourcing surprise lives in the gap between those two.
Understanding who you are buying from changes the conversation. A manufacturer that knits, dyes and finishes in one location can adjust the loop structure when a sample is slightly too stiff. A trading wholesaler buys finished rolls and passes them on, which is efficient for standard items and slow for anything that needs a change.
Keqiao is a useful place to be on either side of that line. The district hosts one of the largest concentrations of knitting, dyeing and finishing capacity in the world, plus the China Textile City market, so a buyer can move from a yarn conversation to a finished-roll inspection within a few kilometres. For a programme that needs several structures in matching shades, that proximity shortens the loop between the first sample and the approved bulk.
A realistic sampling sequence looks like this:
Minimum order quantities depend far more on colour than on structure. Knitting a few hundred kilos of a natural or white fabric is straightforward; dyeing it in a custom shade requires a minimum dye bath, so a 300 kg order spread across five colours can be less practical than a 500 kg order in two. Buyers planning a multi-colour capsule should raise the colour count early, since that is the number that usually drives the MOQ, not the total quantity.
Lead time follows the same logic. Greige knitting is fast; dyeing and finishing are the queue. Allow time for lab dip approval, bulk dyeing, finishing, inspection and shipping, and build in a buffer for the second wash test that many brands skip and later regret.
Colour count, not total volume, usually sets the minimum order; plan the palette before the purchase order, not after.
There is no single best fabric, but there is a best default for each category. For compression leggings and support tops, a nylon/spandex or polyester/spandex interlock in the 220–280 GSM range gives the most reliable combination of support, opacity and recovery. For training tees, a 130–170 GSM polyester jersey or a polyester/spandex single jersey wicks and dries fastest. For warm-up and lifestyle pieces, an air layer, waffle or french terry carries the structure better than a thin jersey ever will.
Nylon generally wins on hand feel, abrasion resistance and compression recovery, which matters on the inner thigh and at the knee. Polyester wins on cost, drying speed and colourfastness in bright shades. Many commercial leggings use a nylon/spandex interlock for the body and a polyester mesh for ventilation panels, accepting a small shrinkage mismatch in exchange for breathability.
As a working guide: single jersey for tees sits around 130–180 GSM, rib for bands and fitted tops around 200–260 GSM, interlock for leggings and bras around 220–280 GSM, and air layer, terry and fleece for warm-up pieces around 240–320 GSM. Mesh panels are typically lighter, in the 90–150 GSM range. These are starting points, and opacity under stretch should always be checked at the actual colour rather than assumed from the weight.
Compression garments commonly use 18–22%, soft leggings and yoga pants 10–15%, fitted tops 5–10%, and joggers or hoodies 3–6%. Above roughly 20%, the fabric becomes harder to heat-set and more sensitive in dyeing, so the percentage should be tied to a recovery target rather than chosen as a round number.
Both are double-knit structures with two sets of loops, but interlock is specifically a balanced structure in which both faces are identical and the fabric is stable on both sides. Double jersey is a broader family that includes interlock alongside structures with a different face and back. In practice, buyers use "interlock" when they mean a smooth, stable, two-faced fabric for leggings, and "double knit" when the faces differ or the structure is heavier.
Usually yes for sampling and counter samples, and often for a first production run in a stock or natural colour. Custom-dyed small lots are the constraint, because a dye bath has a minimum practical volume. A workable approach is to trial the structure in an existing shade, confirm the fit and the wash performance, and then commit the custom colour at production volume.
Cut a strip of known width, mark a fixed gauge length, extend to the agreed elongation, hold briefly and release. Measure the growth after a set relaxation time, and repeat for several cycles. Record the direction of the test, because course and wale behave differently, and run the same test on the sewn seam if the garment relies on the seam for support.
No. Opacity comes from cover factor, yarn count, gauge and colour as much as from weight. A tightly knitted 200 GSM interlock in a dark shade can be more opaque than a loosely knitted 260 GSM fabric in a pale peach. Always confirm opacity on the actual colour, stretched to the tension the garment will see in wear.
Every question above resolves to the same habit: define the garment's job first, then let the structure, blend and specification follow from it.
Athletic apparel fabric rewards specificity. The difference between a programme that runs smoothly and one that burns three sample rounds is rarely a secret material; it is whether the buyer arrived with a garment type, a weight target, a stretch and recovery figure, a colour plan and a price ceiling, and whether the mill could answer with a structure rather than a promise.
Keqiao-based knitting works well for exactly this kind of conversation, because the structures, yarns and finishing capacity sit close together. A programme can move from a nylon/spandex interlock for the compression story to a polyester mesh for ventilation and an air layer for the warm-up piece, with the shades aligned across all three.
Continue with these pages
Bring the garment, the numbers and the colour plan to the first conversation, and the fabric decision becomes a technical one instead of a guessing game.





