Grandma Knitty Home
Knitty: little purls of wisdom
letter from the editorfeatured articlesKnitty's fabulous pattern selectionarchive of back issuestell us what you think of KnittyKnitty's favorite linkshelp knitty keep on keepin' onknitty's virtual sNbjoin the Knitty notifylistknitty's tiny little shopping malltake home something Knitty


the Knitty FAQ

submission guidelines for designers and writers
the obligatory legal statement
the rabbit

© Knitty 2002-2006. All rights reserved. Unauthorized reproduction prohibited. This means you.


Cool stuff! Techniques with Theresa Thinking beyond the pattern Watch this space
Charts are our friends Don't give it away Renovation, transformation Knit by numbers

Ravellings on the knitted sleeve
Part II - Creating a set-in sleeve for a sleeveless body

The whys and wherefores of set-in sleeves

Depending on the intended style of a garment, a drop shoulder or modified drop shoulder is simply less desirable or attractive. For certain styles, such as tailored and close-fitting styles, a set-in sleeve is the most appropriate sleeve to use. To understand why a set-in sleeve succeeds where other styles fail, we need to examine why the set-in sleeve is shaped the way it is.

Parts of the archetypal set-in sleeve

Like any other sleeve, the forearm and upper arm is covered by a main portion that extends from the wrist (if the sleeve is long) to the biceps line, which can be considered the beginning of the underarm. The remainder of the set-in sleeve -- the cap -- extends from the biceps line to the shoulder and has a generally bulbous shape. Following the outline of the cap from the biceps line to the shoulder, you can see that the slope of the outline changes. At the biceps line and shoulder, the slope is pretty much horizontal; between the biceps line and the shoulder, the angle of the cap is first shallow, then steep, then becomes shallow again.

The perimeter of the sleeve cap is sized to fit into the armscye (armhole) of the body, either exactly or with a little excess fabric in the cap that needs to be distributed along the armscye seamline. In a properly fitted garment, particularly in garments sewn from woven cloth, the sleeve cap is shaped differently towards the front and back of the body, and the armscye cut into the body is cut deeper in front than in the back. This difference accommodates the shape of the shoulder blade better than a perfectly symmetric sleeve and armscye. Optionally -- and this was a more common feature in women's clothing until the mid-twentieth century or so -- shaping can be added at the elbow to conform to the arm's natural shape.

However, as you've probably noticed, in hand knitting the typical set-in sleeve does not have quite such sophisticated shaping. The top of the sleeve is finished with a horizontal edge (the final bind-off row). Most set-in sleeves are completely symmetric; there is no elbow shaping, and the back and front halves of the sleeve cap are identical. And since hand-knit fabric usually stretches, we can take advantage of a small degree of forgiveness in the garment shape, and skip the step of precisely tailoring sleeve and armscye. In fact, some set-in sleeve caps are shaped with a constant slope. Normally, hand knit set-in sleeve tops are still sufficiently loose fitting that the lack of asymmetry goes unnoticed.

Why does a set-in sleeve work?

The key to a properly fitting garment is that it has enough ease to allow you to move as you wish. In a loose-fitting garment, that ease is all over the place; there's lots of room in the tubes that act as sleeves, and lots of room in the tube that serves as a body, as well. But in a closer-fitting top, the ease has to be distributed strategically to provide space without excess fabric. In a set-in sleeve, the key is in the cap.

Have you ever tried on a top (either purchased or laboriously knitted by you) and felt a "binding" sensation under the arm? Did the shoulders feel tight? This might have been due to insufficient wearing ease in the shoulder area. For example, the distance from your neck to your elbow, over your shoulder, is shorter when your muscles contract and your arm is extended straight out from your side. When your shoulder is relaxed, with your arm hanging down in the usual position, the distance over your shoulder from your neck to your elbow is longer. The difference in distances isn't large, but it contributes to the binding feeling of a too-tight armscye and sleeve. A garment with sleeves needs to accommodate these little changes in body dimension as the wearer moves his or her arms.

In a really loose-fitting garment, like a drop-shoulder sweater, there's so much space between the garment and body that there's no need to tweak the sleeve/shoulder area to make this accommodation. But in a tight-fitting garment, something's got to give. If there isn't enough fabric to fit or stretch around the body, the result is that tight, binding feeling.

If a typical modified-drop-shoulder pullover is designed to be close-fitting, you may run into that binding problem, particularly with thicker knit fabrics. If you design the pullover to fit your body in the pullover's "relaxed" position -- that is, with the sleeves extending at right angles from the body -- it may fit you perfectly as long as you pose like a woolly board. But when you lower your arms, you'd find you need more space for your shoulders. The obvious place to build in that extra space is, of course, the upper part of the sleeve -- the sleeve cap. If the pullover was not so close fitting, then that binding may not occur at all, meaning there is sufficient space for your shoulders. This suggests that a closer-fitting sleeve needs a deeper sleeve cap.

Also, compare the way you typically stand and move with the typical shape of a drop-shoulder or modified-drop-shoulder garment. Normally, your hands are down by your sides, or at least, your upper arm is not extended at right angles to your body. This suggests that a better-fitting garment would reflect the relative position of your arms to your body in real life, something far less than a right angle to the body. Also, as mentioned earlier, in a drop or modified-drop-shoulder garment (assuming the fit is not close or tight), when the arms are lowered, excess fabric builds up in the underarm area. This means that, rather than having a sleeve cap that is perpendicular to the armscye, the sleeve will fit better if the cap meets the body at an angle smaller than a right angle. This suggests that a closer-fitting sleeve cap needs to be angled, rather than boxy.

Finally, the practicality of fitting square pegs in round holes should be considered. If a sleeve is designed to have a deep, angled cap, it would probably look like this:

While the total perimeter of the sleeve cap is still equal to the perimeter of the armscye, that bind-off distance across the top, unlike the drop or modified-drop-shoulder sleeve, is no longer equal to the vertical drop of the armscye. Those two corners in the sleeve cap fabric would have to be sewn somewhere to the vertical portion of the armscye -- not an easy thing to do, and not particularly attractive. This suggests that to have a smoothly fitting sleeve, the sleeve cap should be gently curved from underarm to shoulder point, rather than abruptly angled.

And of course, the human figure isn't rectangular; it's curvy. The imaginary line demarcating the body-arm boundary isn't a box, so why knit one? A more refined sleeve and body, then, would look something like this:

And we've arrived at the familiar set-in sleeve shape.

Designing a set-in sleeve to fit a sleeveless top

A little more theory

By merely inspecting the respective shapes of the armscye and sleeve cap, it's easy to see that the method of calculating set-in sleeve shaping must be more involved than a simple drop shoulder or modified-drop shoulder. The geometry is more advanced; the curved or sloped sides of the sleeve cap must fit into the straight edge of the armscye, and some fixed number of stitches at the very end must be bound off without creating corners of fabric that are too troublesome to sew into the armscye. As for the main part of the sleeve itself, sufficient ease must be built into the sleeve, but not so much or so little that the sleeve looks disproportionate to the garment body; and the length, including the depth of the cap, must be chosen to hit the wrist (in the case of a long sleeve) at just the right point.

In fact, different designers and pattern drafters have different techniques for calculating set-in sleeves to fit a given armscye, and follow different rules of thumb for:

  • determining the height of a sleeve cap;

  • deciding whether the sleeve cap shaping should have a constant slope or a varying slope;

  • deciding how many stitches to bind off at the top of the sleeve cap.

These different rules of thumb vary according to what numbers are fixed or calculated first. Some designers establish the height of the sleeve cap first, while others fix the number of stitches to be bound off before considering the height or the slope. As a general rule, however, the shaping and fit of a plain set-in sleeve is dictated by the shaping and fit of the body: if the body is close fitting, the sleeve is close fitting and the sleeve cap is proportionately deeper; if the body is loose fitting, the sleeve is loose fitting and the sleeve cap is shallower. Exceptions are made, of course, for sleeve style variations, such as pleated or gathered caps.

Other matters, such as how the biceps measurement, circumference of the wrist, length of the sleeve, or location of the underarm are determined, are sometimes handled differently by various designers, but the end result is the same.

Begin with taking measurements

The sleeve design is premised on certain dimensions:

Working from an existing sleeveless top pattern

If you are starting from a sleeveless top pattern, choose a design with armholes that approximate the shaping of a set-in sleeve top:

  • an armscye beginning within an inch or two of the wearer's underarm (more, if the garment is loose fitting) curving as it emerges from the underarm until it forms a more-or-less vertical line over the shoulder; and

  • shoulder coverage extending from the shoulder line, or near the shoulder line, towards the neck (the amount of shoulder coverage depends on the width of the neckline).

A halter top does not have the same amount of shoulder coverage, and the armscye is often always angled towards the neck. Strappy and strapless tops don't have an armscye at all, although if the straps of a strappy top are positioned near the shoulder line, you can treat the underarm-strap assembly as an armscye for a set-in sleeve. (You may find that skinny straps positioned so close to the shoulder will tend to slip off, making them useless for holding up a set-in sleeve. In that case, consider widening the straps to cover more of the shoulder and adding means to fix the garment straps to the straps of your underpinnings.)

a) the armscye depth on the garment body;

b) the armscye perimeter on the garment body;

c) the length of the arm from armscye to wrist (if the sleeve is meant to be that long);

d) the width of the sleeve at the biceps line; and

e) the circumference at the wrist (or the circumference at the lower edge of the sleeve, if the sleeve is not wrist length).

With the exception of the armscye dimensions a and b, these measurements are based on actual body measurements, plus or minus ease. The amount of ease depends on the overall style and fit of the garment. Often, the body fit and design has already been established, so the amount of sleeve ease will be determined based on the body ease.

Let's assume you are starting from an existing sleeveless top pattern. From its schematics, or from counting the rows in the armscye, you can work out the depth of the armscye, a. Make a note of that value. Although measuring the perimeter of the top's armscye is likely not possible without some trigonometric number crunching or drawing the pattern schematic to scale, you can at least make a note of the number of stitches bound off at the beginning of the armscye on one side and arrive at a rough estimate of the total perimeter of the armscye. You can also verify the fit of the armscye by comparing the measured depth a with your actual body measurement. In a standard fitting garment, the armscye ends about 1 inch below the actual armpit--more, if the yarn is bulky or you intend to wear loose shirts underneath the garment. Also make a note of the distance between the actual armpit and the lower edge of the armscye. This value will be used to determine the length of the sleeve.

A word of caution about working from an existing sleeveless top pattern: the set-in sleeve seam is typically expected to lie across the shoulder point, and the armscye perimeter is generally loose enough to allow for comfortable wear and freedom of movement. However, sleeveless tops are often cut to expose a little more shoulder, and a little less armpit. This means that you may need to adjust your existing pattern to decrease fewer stitches at the armscye shaping so that the upper back width is equal to your shoulder-to-shoulder measurement, and to drop the lower edge of the armscye so that there will be sufficient ease to fit a sleeve.


The length of the sleeve from armscye to wrist, c, is not the same as the total length of the sleeve from the top of the sleeve cap to wrist. This particular measurement is measured essentially from the garment's "armpit" to the wrist. For this value, measure from the wrist, over a slightly bent elbow, and up to your actual armpit, and then subtract the distance you measured above, between the actual armpit and the lower edge of the garment armscye.

Also from the existing pattern schematic for the top, make a note of the total ease around the bust or chest (that is, the difference between the actual body measurement and the garment measurement at the widest point at the chest. For a plain set-in sleeve, the width of the sleeve at the biceps line, d, will be the circumference of the upper arm at its widest point plus about 50-100% of the body ease. This 50-100% of the body ease, in a "standard fitting" garment, is typically 1 or 2 inches. For a styled sleeve, like a puffed or gathered sleeve, the width at the biceps line may be far greater than this amount.

Ease varies over time

If you consult a number of reference books, the amount of ease in the body will fluctuate according to the silhouettes in vogue at the time the book was published.

Sweaters from the 1980s and early 1990s, for example, had loose, easy-fitting styles, and consequently an "average" fitting sweater would have at least 4 inches of ease. In the late 90s and early 00s, a fashionista would have chosen a closer fit as an "average" fitting sweater.

This means, when reading garment descriptions, you should be aware that a "classic" or "standard" fit may not be what you consider to be your classic or standard. In this article, "classic" is assumed to be 2 inches of ease.

Finally, the wrist circumference e is generally taken to be either the actual wrist measurement plus 1 inch, or a measurement sufficient to pass over a closed fist.

Sketch your schematic and write in the measurements you know

Sketch a large schematic of the sleeve, approximating the shape of the sleeve cap. If you are using proportional graph paper, don't bother graphing the entire sleeve to scale--the main part of the sleeve is quite straightforward. Save the graph paper for sketching the cap alone. Make a note of your stitch and row gauge (number of stitches and rows per inch) on your schematic, as well as some of the measurements you've taken so far: the length of the sleeve c, the width of the sleeve at the biceps line d, and the circumference at the wrist e.

Convert the numbers you have to stitches and rows

With your knitting gauge, convert the measurements you've written on the schematic to stitch and row counts, adding a selvedge stitch or two. At this point, you'll be able to work out the increases for the main part of the sleeve, just as the sleeve shaping was calculated in last issue's article. In this article, we'll skip the addition of a cuff, and just work straight up from the wrist.

Also, at this point, at the start of the sleeve cap, make a note of how many stitches to bind off the first two rows of the sleeve cap. One of the more constant rules of thumb in knitwear design is that the number of stitches bound off at the beginning of the garment armscye equals the number of stitches bound off at the beginning of the first two rows of the set-in sleeve cap.

If you are using proportional graph paper, now's the time to mark off the beginning of the sleeve cap, and the first two rows of the cap shaping.

Design the sleeve cap to fit the armscye

Next, the sleeve cap shaping needs to be determined. This is where the design technique can vary. Some designers will proceed in the following order:

  1. Fix the height of the sleeve cap.

  2. Fix the number of stitches for the final bind-off of the sleeve cap.

  3. Decide how to work the decreases to fit between the initial bind-offs and the final bind-off, so that the sleeve cap fits the perimeter of the armscye.

This particular sequence is likely to be followed if you choose to follow the rule of thumb that a sleeve cap should have a particular height. In her book, Designing Knitwear, Deborah Newton recommends that a sleeve cap for a "classic"-fitting garment have a height equal to about two-thirds of the depth of the armscye, a; for a loose-fitting garment, this height should be less, and for a closer-fitting garment, the height should be greater (note that this fits in with the theory discussed above, that a closer-fitting sleeve needs a deeper sleeve cap).

The final bind-off of the sleeve cap is often fixed between 2 and 6 inches; less for smaller sizes, more for larger sizes, and less for closer-fitting garments, more for looser-fitting garments. If the corners produced by the bind-off are to be softened by a few sets of bind-off rows, choose a narrower final bind-off.

When the rest of the sleeve cap is drafted, the aim is to have a sleeve cap perimeter that is at least the same measurement as the armscye perimeter. The sleeve cap perimeter may be about 1 inch greater than the armscye perimeter without causing any difficulties in sewing up, so don't worry about being precise; the extra fabric can be eased in invisibly. Just don't design a sleeve cap that is too small to fit within the armscye perimeter.

There is obviously some uncertainty or at least some recursive tweaking done with sleeve cap decreases to get the sleeve to fit the armscye properly. This method is most easily accomplished if graph paper ruled to match the knitting gauge is used -- the outline of the sleeve cap can be sketched, then measured with the edge of a measuring tape to confirm the sleeve cap perimeter measurement. The outline can then be converted to actual knitted stitches by tracing over the closest gridlines.

Other designers might follow this order:

  1. Fix the number of stitches for the final bind-off of the sleeve cap.

  2. Decide how to work the decreases to fit between the initial bind-offs and the final bind-off so that the sleeve cap fits the perimeter of the armscye.

In this case, the final height of the sleeve cap is an unknown variable, and the designer doesn't really care what it is, as long as the sleeve cap perimeter fits and looks like a sleeve cap. The final bind-off rows are chosen in accordance with the same rules of thumb as the first method.

If this second method were to be applied to graph paper, then it would probably wind up being implemented the same way as the first method -- a cap height would have to be chosen first, otherwise the graphing step would not make much sense. Designers who implement this second method likely do not use proportional graph paper at all, but rather use trigonometry to design a sleeve cap with the desired bell or angular shape: since the perimeter of the sleeve cap provided by the initial and final bind-offs is known, the remaining perimeter must be provided by a series of decrease rows. The number of required decrease rows is determined using a little trigonometry -- a little more trigonometry is required if the sleeve cap is to have a bell shape, with varying rates of decreases.

Still another order of operations is:

  1. After the first two bind-off rows in the sleeve cap, decrease according to a set formula until the perimeter of the sleeve cap fits the perimeter of the armscye.

  2. Bind off the remaining stitches.

This method is effectively the "live" method of knitting a sleeve cap -- rather than engage in calculations or graphs beforehand, once the first two bind-off rows are worked, the knitter simply wings it until the sleeve cap looks like it will fit. The height of the sleeve cap is unknown, and so is the width of the final bind-off. However, there is less work to do in computing the decreases since it follows a set formula, typically decreasing one stitch at either end of every other row. This means the shape of the sleeve cap is a simple trapezoid, rather than a bell curve. This method is safer for looser-fitting garments, since the fit to the body is imprecise and the width of the final bind-off could be relatively wide.

You can check the length of the sleeve by measuring your total "wingspan" -- that is, the measurement from a first wrist, over a first slightly bent elbow, up the arm, over the shoulder, across the back neck, and down the other side following the same path. Your total wingspan should be equal to the total sleeve length of the first sleeve (including cap), plus the upper back width, plus the total sleeve length of the second sleeve.

If desired, write out the schematic or graph as knitting instructions

If the sleeve cap was designed using the first or second method, then this will be fairly straightforward. If you used the third, "live" method, then you will have to write out your instructions as you knit.


Final notes on set-in sleeve construction

First, if you tend to knit your sleeves before the body, this is one situation where it is probably better to knit the body first. It's easier to test the fit of the sleeve cap in the armscye if the armscye is completed first; if it doesn't fit, only some or all of the sleeve cap needs to be ripped out and reknit. If the body is already constructed, you can even test the fit of the sleeve cap and estimate the shaping of the remainder of the cap as you are knitting. If the sleeve was completed first and it turned out that the armscye did not fit the sleeve cap, you would be faced with the choice of ripping back the sleeve cap, or ripping back the upper body.

On the other hand, if you are married to the method of using your sleeves as a giant gauge swatch, just knit the sleeves up to the beginning of the sleeve cap. Let them wait on waste yarn or a stitch holder until the body is completed.

Secondly, for your first attempt at designing set-in sleeves, it's easier to work in flat pieces, knit from the wrist up. It is possible to knit set-in sleeves from the armscye downwards by using short rows to shape the sleeve cap, and it's a rather clever strategy. However, it's easier to understand this technique (for example, as described in Barbara Walker's Knitting from the Top) once you understand how the flat pieces fit together.

You might also wonder about selvedge stitches: should they be included in your calculations in the sleeve cap? Generally, if you have built in a single-stitch or half-stitch selvedge into your garment bust measurement and sleeve width, you can pretty much ignore the effect of the selvedge when you are attempting to fit the sleeve cap to the armscye. The overall fit along that seamline will not be affected.

Finally, what's to stop you from simply making up a sleeve cap as you go along, if you're designing from scratch? Is it necessary to engage in all the computation and sketching? No, if you don't mind knitting and ripping repeatedly to get a sleeve cap to fit, by all means, do it. However, if you're trying to design for a range of sizes, you're unlikely to want to knit a series of sleeves and you'd probably save time by hauling out the calculator and graph paper.



These books contain information on drafting set-in sleeves for knitting:

In print:

Newton, Deborah, Designing Knitwear (Taunton, 1992).
Vogue Knitting: The Ultimate Knitting Book (Random House, 1989). An updated edition is now in print.

Out of print:

Duncan, Ida Riley, The Complete Book of Progressive Knitting (Liveright, 1940); Knit to Fit (Liveright, 1970).
Elalouf, Sion, The Knitting Architect (Knitting Fever Inc., 1982) (cerlox-bound booklet)
Michelson, Carmen, and Mary-Ann Davis, The Knitter's Guide to Sweater Design (Interweave Press, 1989).



Jenna has a bias in favor of set-in sleeves.