Cashmere Pilling: Why It Happens and 3 B2B Manufacturing Choices That Cut It by 70%

Pilling is the single most common cashmere complaint a B2B brand receives, and it is almost never a fiber-quality problem. It is a yarn-construction and finishing decision made at the mill.

This explains the actual mechanism — how a fiber knot forms and why cashmere’s short staple makes it more likely — the three construction choices that reduce pilling by roughly 70% without changing the hand feel, the finishing step most mills skip, and the test method you should specify in your purchase order so you can verify rather than hope.

What a pill actually is

A pill is a knot. It forms when a fiber that has migrated to the surface of the fabric is then rubbed — by a shoulder strap, a seatbelt, a sleeve, or a laundry cycle — and gets caught on an adjacent fiber and twisted into a tight ball.

For pilling to happen you need three things at once:

1. A fiber on the surface. Looser constructions push more fiber to the surface.
2. A fiber long enough to wrap. Short fibers wrap easily; long ones stay anchored.
3. An abrasive contact. Friction is the trigger, not time.

This is why pilling appears in the first few wears rather than after a year. It is a surface phenomenon, and once the surface fibers are consumed the fabric stops pilling — which is why “it pills for a month then stops” is a real and common pattern.

Why cashmere pills more than merino

Cashmere’s staple length runs 30-80mm. Merino runs 70-120mm. Cashmere has roughly half the fiber length, and a short fiber has fewer points of contact inside the yarn, so it migrates to the surface more readily and wraps more easily once it gets there.

This is a structural property of the fiber, not a defect. It is also why cashmere pilling is much more visible than merino pilling at the same pill size — cashmere pills are the same color family as the fabric and read as a visible fuzz, while a merino pill in a dark color often matches the ground and is invisible until it lightens.

The 3 construction choices that cut pilling by ~70%

These are decisions made at the yarn and knitting stage. None of them change the hand feel materially.

1. Raise the yarn twist

Twist is what holds a yarn together. A loosely-twisted cashmere yarn has fewer fiber-to-fiber contacts inside the twist, so fibers migrate more easily.

The change: specify twist at or above the upper end of the normal band for the count.

In your spec: twist Z, 520-560 T/m for 2/26 Nm worstled

Typical cashmere worstled yarn runs 480-520 T/m. Pushing to 520-560 adds roughly 8-12% more twist, which measurably reduces surface migration at no cost to hand feel — the fiber is still fully soft, it is just held more tightly.

The tradeoff is a slightly harder yarn that knits with marginally higher needle wear. For a hand-knit or fine-gauge product this is not worth it; for a 7GG-12GG machine-knit it usually is.

2. Use a longer staple within the micron band

Two cashmere lots at 15.5μm can differ in staple length by 20mm. The lot with the longer staple pills less at identical diameter and identical price.

In your spec: staple length 42-58mm mean, in addition to the micron band

This is the cheapest available pilling improvement because it costs the mill nothing — it is a selection decision at the sorting stage, not a processing decision. Ask for it explicitly; it is not a default.

3. Knit tighter

Fabric density controls how much fiber is exposed. A loosely-knitted cashmere fabric pushes more fiber to the surface than a tight one.

In your spec: knit density within 8% of the specified stitches-per-cm, verified on the first delivery

The practical version: sample the fabric on the machine before bulk, measure it, and hold the bulk to that measurement rather than to a tolerance band that allows 15% variation.

The finishing step most mills skip

The step: wet finishing / fulling followed by a brushing pass in the opposite direction.

Wet finishing relaxes the yarn and sets the knit. Brushing raises the surface nap. But brushing in the same direction as the nap lays fibers flat in one orientation, which makes them easier to catch. Brushing against the nap raises the fibers in multiple directions, which distributes the abrasion across many fibers instead of concentrating it on a few.

This is a finishing-line setting, not a material change. It costs the mill almost nothing and it is not in most purchase orders because most buyers do not know to ask for it.

The test you should put in the purchase order

Do not specify “anti-pilling” as a requirement — it is not a measurable term. Specify the test and the grade.

In your spec: pilling grade 4-5 minimum after 1,000 Martindale rubs per ISO 12945-2, tested on the finished fabric after wet finishing.

ISO 12945-2 is the standard. The 1,000-rub level corresponds to roughly the first several months of actual wear for a sweater. Grade 4-5 at 1,000 rubs is a demanding but achievable target for a well-constructed cashmere.

If a mill tells you it cannot hit grade 4-5 at 1,000 rubs, that is useful information — it means either the fiber is at the coarse end of the band or the construction is at the loose end, and you now know which.

What actually does not work

Anti-pilling agents. Chemical treatments that reduce fiber migration exist and they work for a few washes. They also change the hand feel, they are not reversible, and they do not survive an audit of what the garment is made of. If a brand has an OEKO-TEX claim, the treatment has to be inside the certified scope, which limits what can be used.

Blending in a longer fiber as a fix. A cashmere-merino blend pills less, but it pills less because it is less cashmere, not because it is better cashmere. If the product is positioned as cashmere, that is not a construction improvement.

Telling the customer to de-pill with a comb. That is a customer-experience measure, not a quality measure. It can be part of the insert card, but it does not fix the product.

The mill-side view

At our Ordos facility the default specification for a commercial cashmere line is: 15.5-16.5μm, staple length 45mm mean, twist 520 T/m, knit density held to ±6% of the approved sample, ISO 12945-2 grade 4 minimum at 1,000 rubs. That combination is not exotic — it is roughly 70% better than the looser end of what the market accepts, and it costs us about 3% in yarn consumption and nothing in hand feel.

If you want to know what your current construction scores, send us a fabric sample and we will run the test and report the actual grade.

For the construction questions that come up alongside pilling — how the yarn itself is built — see our guide to cashmere yarn counts. For the finishing-side quality controls that sit next to this, see our AQL final inspection guide.