Naoshu

What is SI10?

SI10 (Naoshu) is a point on the back of the shoulder, in the hollow just below the outer end of the shoulder blade spine.

It is the tenth point on the Small Intestine channel. The name Naoshu is usually translated as "upper arm point", which matches how tradition used it: for a sore, stiff or heavy shoulder and a weak arm.

In classical texts SI10 is described as a meeting point of the Small Intestine channel with two of the extraordinary vessels, the Yang Linking and Yang Motility vessels. That gave it a reputation beyond simple local use, and most modern point charts still list it among the standard shoulder points alongside SI9 and LI15.

Here is the honest verdict on the evidence. There are no published human trials testing finger pressure on SI10 by itself for any condition. Where SI10 appears in research, it is almost always one name in a longer needling prescription for shoulder pain or post-stroke arm weakness, so nothing can be credited to this point alone. The wider acupressure literature is mostly small, unblinded and short-term, and that limitation applies here.

One thing worth saying clearly, once, for the whole page: a long traditional pedigree is not evidence that a point works, and it is not evidence that pressing it is safe for you. Those are two separate questions, and tradition answers neither. The safety advice further down this page comes from anatomy and ordinary clinical common sense, not from the classics.

What SI10 reasonably is: a comfortable, easy-to-reach spot for self-massage on a tight posterior shoulder. Firm thumb pressure on tense muscle often feels good and may ease stiffness for a while. That is a modest, believable claim, and it is the one this page supports.

Where is SI10 and how do you find it?

SI10 sits on the back of the shoulder, in the soft hollow directly below the bony ridge of the shoulder blade, near its outer end.

• Start at the back armpit fold. Let your arm hang at your side. Reach across with your opposite hand and find the crease where the back of your armpit meets your arm. That fold is your starting mark.
• Move straight up. From the top of that fold, travel upward about one thumb-width, roughly 2 to 2.5 cm. In traditional measurement this is 1 cun above the posterior axillary crease.
• Find the bony shelf. Keep moving up until your fingertip bumps the underside of a firm horizontal ridge of bone. That ridge is the spine of the shoulder blade, running out toward the point of the shoulder.
• Settle into the hollow. SI10 is in the soft depression just below that bony shelf, not on it. You should feel muscle under your finger, not hard bone.
• Check with a small movement. Lift your arm out to the side a few centimetres and lower it. The tissue under your finger should shift and tighten slightly, which confirms you are on muscle in the right hollow.

A simple reference: SI10 sits above SI9 and behind the point of the shoulder, on the same vertical line as the back armpit fold. If you have landed on top of the bony ridge, you are too high. If you are sliding into the armpit hollow itself, you are too low and too far forward.

Most people find SI10 easier to reach with the opposite hand coming over the front of the chest, rather than reaching behind. If your shoulder is too stiff for either, a helper can find it for you, or you can rest the spot against a soft ball on a wall.

How do you apply pressure to SI10?

1. Get the shoulder relaxed. Sit upright and let the arm hang loose, or rest the hand in your lap. A shoulder held up near the ear is hard to press usefully.
2. Place the thumb or two fingers. Reach across your chest with the opposite hand and settle your thumb pad into the hollow below the bony ridge. Use the flat pad, not the nail edge.
3. Press in slowly. Lean the pressure inward and slightly upward, toward the underside of the bony shelf. Build over about five seconds rather than jabbing.
4. Aim for firm, not fierce. On a scale where ten is unbearable, stay around four to six. It should feel like a satisfying ache that you could hold a conversation through.
5. Hold and breathe. Keep steady pressure for 30 to 60 seconds, breathing normally. Small slow circles, about one per second, are fine if that feels better than a static hold.
6. Release and repeat. Ease off over a few seconds, rest for half a minute, then repeat two or three times. Do the same on the other shoulder if both are tight.
7. Use a ball if you cannot reach. Stand with a soft massage ball between the spot and a wall, and lean gently. Control the pressure by how much weight you put through it.
8. Finish with movement. Roll the shoulder slowly backwards a few times. Pressure plus gentle movement usually feels better than pressure alone.

How often: once or twice a day is plenty for a tight shoulder, and a few minutes each time is enough. There is no benefit in grinding away for twenty minutes.

Stop straight away if you feel sharp pain, pins and needles down the arm, numbness, or any sense that the shoulder is about to give way. Mild tenderness for an hour afterwards is normal; bruising means you pressed too hard and should use lighter pressure next time.

What is SI10 traditionally used for?

• Shoulder pain and stiffness. This is the main classical use, and the most plausible one. Firm pressure on tight muscle around the back of the shoulder often eases stiffness short-term, but no trial has tested SI10 acupressure on its own, so treat it as comfort rather than treatment.
• Aching or heavy upper arm. The name Naoshu points to the upper arm, and old texts list it for an arm that feels weak or hard to lift. Human evidence for this specific effect is absent; what you get is local muscle relief, which may make the arm feel lighter for a while.
• Limited shoulder movement. Tradition used SI10 when the arm could not be raised easily, including the pattern now often labelled frozen shoulder. Acupuncture research on frozen shoulder is mixed and mostly involves needling several points at once, so nothing can be attributed to SI10 alone.
• Post-stroke arm weakness. SI10 appears in some Chinese needling protocols for a weak arm after stroke. Those studies are small, rarely blinded, and used needles rather than fingers. Self-applied pressure should never replace stroke rehabilitation.
• Desk and screen neck-shoulder tension. A modern, non-classical use. The muscle under SI10 is a common tender spot in people who sit at keyboards. Self-massage here is reasonable for comfort; no study has measured it.
• Swelling around the shoulder. Some classical sources mention swelling and "wind" in the shoulder region. There is no modern evidence behind this, and new or unexplained swelling needs a doctor, not pressure.

Taken together: SI10 has a clear traditional identity as a local shoulder point, and essentially no point-specific human evidence. The reasonable expectation is temporary easing of muscular tightness, similar to any other self-massage on a tight spot.

Are there tips for using SI10 effectively?

• Warm the area first. A few minutes under a warm shower, or a heat pack on the shoulder, softens the muscle and makes the same pressure feel less sharp.
• Press below the ridge, not on it. If your thumb is on something hard, slide down a centimetre until you feel soft tissue. SI10 sits in the hollow, and pressing bone achieves nothing but soreness.
• Let the arm hang heavy. Tensed muscle resists pressure. Drop the shoulder, let the arm dangle or rest in your lap, and the tissue will give under your thumb much more easily.
• Use a ball against the wall for a long hold. Reaching across your chest gets tiring within a minute. A soft ball between the spot and a wall lets you hold comfortable pressure for longer and control it with your body weight.
• Pair it with shoulder movement. Press for a minute, release, then do slow shoulder rolls or reach the arm across your chest. Pressure plus movement tends to help stiffness more than pressure alone.
• Work both sides. Shoulders usually tighten as a pair, especially from desk work or carrying a bag. Doing one side only often leaves things feeling lopsided.
• Keep it short and regular. Two or three minutes once or twice a day beats one long, hard session a week. Over-pressing causes bruising and fresh soreness.
• Track honestly for two weeks. Note your stiffness out of ten before and after. If nothing has shifted in a fortnight, the point is not the answer for you, and that is useful information.
• Fix the cause too. Screen height, bag strap, sleeping position and shoulder strength all matter more than any point. Use SI10 as relief while you sort those out.

One practical note: if reaching across your chest hurts the shoulder you are trying to help, stop and ask someone else to find the spot for you.

What do people get wrong with SI10?

• Expecting a cure. The single biggest error. SI10 is not going to fix a torn rotator cuff, a frozen shoulder or nerve damage. At best it eases muscular tightness for a while.
• Pressing on bone. Landing on the hard ridge of the shoulder blade or on the point of the shoulder, instead of the soft hollow just underneath it. Bone pressure is sore and pointless.
• Digging into the armpit. Sliding too low and forward puts pressure into the armpit, where nerves and blood vessels and lymph nodes sit close to the surface. That is the wrong place and can cause tingling down the arm.
• Confusing it with SI9 or LI15. SI9 sits lower, about a thumb-width above the back armpit fold but below the ridge, and LI15 is in the dip at the front-outer shoulder. SI10 is the one tucked directly under the bony ridge.
• Pressing far too hard. Chasing a dramatic ache leads to bruising and a shoulder that hurts more the next day. Firm and tolerable is the target, not eye-watering.
• Working through sharp or shooting pain. Tenderness is fine; sharp pain, numbness or electric sensations down the arm mean stop. Those are nerve signals, not a sign it is working.
• Using it instead of getting a diagnosis. Shoulder pain that wakes you at night, follows a fall, or comes with arm weakness needs assessment. Months of self-massage can delay a diagnosis that matters.
• Pressing a shoulder that has just been operated on. Fresh scars, recent surgery and new hardware are off limits until your surgeon clears hands-on work.
• Doing it once and giving up. If you are going to try it, give it a fair, gentle two weeks and judge it on your own notes.

What does the research on SI10 actually say?

There is no published human research testing SI10 on its own, by finger pressure or any other method. That is the plain answer, and it is the honest starting point for this page.

Where SI10 does appear in the literature, it is as one name inside a longer list of points in acupuncture trials for shoulder pain, frozen shoulder, or arm weakness after stroke. Those studies used needles, not fingers, and they treated several points together in one session. Even when such a trial reports a benefit, there is no way to tell how much, if any, came from SI10 rather than from the other points, the handling, the expectation of help, or simply time passing.

The broader picture is worth knowing. Acupuncture for shoulder pain has been studied reasonably often, and reviews of that work tend to reach the same cautious conclusion: trials are generally small, poorly blinded, short in follow-up, and inconsistent in which points they use. Some show modest short-term pain relief over no treatment, and the gap narrows against sham or placebo needling. Acupressure specifically — pressure with fingers — has far less research again, and most of it concerns nausea, labour pain and sleep rather than shoulders.

So the honest reading for SI10 is this. The evidence grade for this point and this use is insufficient, not because it has been tested and failed, but because it has barely been tested at all. Its traditional standing is solid: it is in the standard point lists and taught in most curricula as a local shoulder point.

What that means for you is simple. Pressing SI10 is best understood as self-massage of a tender shoulder muscle. That is often pleasant and sometimes helpful, and it is not a treatment for a diagnosed shoulder problem. If anyone cites a study proving SI10 works for a named condition, ask to see it, and check whether the study needled that point alone.