The Trapezius Muscle Myth

The trapezius is a large, triangular shaped muscle that covers the back of the neck, shoulders and thoracic spine. The muscle is usually described as having three sections: the upper, middle, and lower portions. Physical therapists often talk about an imbalance in the strength and activity of the different sections of the trapezius contributing to shoulder problems. Consequently specific exercises are used to isolate the different portions of the muscle.

One question I think deserves asking: are there really three different parts of the trapezius?

Trapezius Anatomy

The trapezius is a large muscle and is undoubtedly involved in many movements. The muscle spans from the base of the skull all the way down the back to the 12th thoracic vertebrae. The muscle inserts on the shoulder girdle at the distal portion of the clavicle, the acromion, and the spine of the scapula.


The imbalance most commonly noted as causing shoulder problems is an increased ratio of upper trapezius activity in relation to lower trapezius activity. Claiming an imbalance implies that the upper and lower portions of the trapezius are capable of contracting with some degree of independence.

In some animals there are three distinct trapezius muscles, but in humans they form one continuous muscle. Since all of the fibers of the trapezius are thought to constitute a single muscle, it’s a valid question to ask if this independence is possible. Can one portion of the trapezius be stronger, more active, and be contracted preferentially from the other portions?

The Trapezius As One Muscle

Anatomy textbooks typically denote the separate roles of the different sections of the muscle as:

  • the upper trapezius elevates and laterally rotates the scapula
  • the middle trapezius retracts the scapular
  • the lower trapezius depresses and medially rotates the scapular

These actions make it seem as if the trapezius is made up of three distinct muscles. There are even some highly questionable manual muscle tests designed to measured the relative strength of each of the sections.

The function of different sections of the trapezius has recently been called into question, particularly in regard to the upper trap elevating the scapula as detailed in this excellent article by Adam Meakins.

Contrary to previous thinking, it’s becoming better established that the upper trapezius is not the main elevator of the shoulder. The fibers of the upper trap are small and badly positioned for this role. As Adam also points out, the trapezius does not work in isolation. Rather, it works in synergy with less superficial muscles.

Scapula Muscles

Because of its multiple insertions and fiber orientations, the trapezius will certainly influence movement in different ways depending on the position of the scapula relative to the spine. There may also be separate motor nerve branches supplying each section. The main nerve supply comes from the spinal accessory nerve. There is some question whether other nerves from the neck also supply the muscle.

I’m curious why anatomists chose to divide the muscle into these sections. It’s hard to believe that different sections of a continuous muscle can be activated separately. I have not seen evidence that the upper, middle, and lower traps function independently.

Several studies have looked at the supposed imbalance in the trapezius, including this one which appeared to find a higher ratio of upper trapezius to lower trapezius activity in subjects with shoulder impingement. These studies are based on the assumption that there are indeed three separate functioning parts of the trapezius.

Most studies on trapezius activity also rely on surface EMG measurements. Eletromyography (EMG) involves places electrodes on the skin overlying the muscle and recording electrical signals. EMG is useful, but it has limitations.

  • One problem with EMG is crosstalk, or picking up signals not just from the muscle being studied, but also the activity of muscles nearby. For example, attempting to measure upper trapezius activity during a shoulder movement might also measure levator scapulae activity.
  • A second problem is that EMG activity can be influenced by the velocity of muscle contraction.

Other factors that can influence EMG readings include:

  • subcutaneous tissue
  • changes in electrode spacing with movement
  • muscle movement relative to the electrodes

Because of these limitation, EMG studies need to be interpreted with caution. What appears to be an imbalance in electrical activity in the trapezius may not actually mean a difference in activation of fibers at either end of the muscle.

Can Exercise Target One Portion of a Muscle?

Assuming the trapezius is in fact one muscle, then fundamental to answering the original question is determining whether parts of a muscle can be preferentially activated. Does one portion of a muscle get activated more than another part for a particular movement?

Different exercises vary the amount of tension and stress on different parts of a muscle. A single muscle can have multiple never suppliers as well. Whether this increases the recruitment of motor units in the part of the muscle best positioned to assist with a certain movement or exercise is not clear. The neuromuscular system is complex, so it’s possible. This is not something that should be assumed however.

This issue is similar to the debate over whether it’s possible to isolate the vastus medialis obliquus (VMO) portion of the quadriceps for knee rehabilitation. The quadriceps muscle group is actually four separate muscles that form a common tendon and work together to extend the knee. It was once thought that a weaker VMO in relation to the vastus lateralis (VL) part of the quadriceps was the cause of patellar tracking issues. A systematic review on the subject that included 20 different studies concluded that the VMO cannot be preferentially activated and strengthened. In other words, the quadriceps can only be strengthened as a whole. It is not possible to target a specific part.

One of the most debated cases of this is with the rectus abdominis. Upper and lower abs. There is a lot of conflicting evidence on whether exercise can target the upper or lower part of the rectus abdominis, and if differences do exist they may be small.

Bodybuilders frequently talk about doing exercises to target the upper chest. Anatomically, this refers to the upper portion of the pectoralis major muscle that attaches to the clavicle. The pectoralis major has innervation by both the medial and lateral pectoral nerves. Contrary to popular belief though, incline bench pressing does not appear to increase activation in the upper portion of the pectoralis major.

The deltoid muscle of the shoulder may be the most relevant to this discussion as it is also commonly said to function in parts, with each part performing a different movement. Anatomically, the deltoid is said to have three heads: anterior, medial, and posterior. The anterior and middle deltoid are innervated by the anterior trunk of the axillary nerve. The posterior deltoid is innervated by the posterior trunk of the axillary nerve. Studies show that there may be different levels of activation in the different regions of the deltoid depending on the shoulder movement being performed. Again, the conclusions that can be drawn from all these studies are subject to the limitations of EMG measurement.

Most importantly, none of the studies show that you can activate one part of a muscle while the rest stays dormant. This is true for the quads, abdominals, pecs, and delts. Regardless of the movement, when a muscle gets used the whole muscle is activated. This holds true even for muscles with separate heads and multiple nerve suppliers. It’s not possible to switch on one portion of a muscle, while switching off another, as is often attempted with lower trapezius strengthening exercise.


It’s unlikely that the lower trap can be activated without activating the upper and middle portions of the muscle as well. When we talk about targeting the lower trapezius, we really mean improving the strength and synergy of the group of muscles that control motion of the shoulder blade.

The trapezius acts to stabilize the shoulder and transfer mechanical tension. It holds the shoulder girdle up like a suspension bridge and controls the position of the shoulder blade when the arm is moving. This stabilization occurs through contraction of the entire muscle, not just a portion. Problems like shoulder impingement should probably not be attributed to a trapezius imbalance or “weak lower traps”. Likewise, an “overactive upper trap” is likely not the cause of neck tension or an elevated shoulder girdle. We can still work on improving scapular depression and rotation, but the explanation for how we do this needs to change. 


  1. says

    Really good information there and a very good reminder that as Mel Siff said, “The body knows of movements, not muscles.” Gray Cook is also good about observing and correcting movement patterns not individual muscles. I tend to sometimes fall into the trap of thinking of a single muscle when really I should stick to the observable movement pattern and quit worrying about which muscle is doing what.


  2. David Garcia says

    Good review, just a quick addition. All your statements make sense from the biomechanical standpoint , howewer, the article by Cools et al and Janda’s work show that the UT is a muscle that has a tendency towards becoming overactive from the Neuro standpoint. That’s where that research points to. Thanks again, David

    • James Speck says

      Hi David, thanks for commenting. The work of Cools and Janda is based on the assumption that the upper and lower trapezius function as separate muscles. The main reasons I wrote this article was to highlight that this assumption may be wrong.

  3. Kris, PT says


    I do think the article poses a valid point. Let’s keep challenging some anatomical and functional anatomical concepts. We really don’t know as much as the text books show us about nerve and muscle anatomy/mechanics, despite how simple the “artistic” representations our anatomy text book lays it out for us. There are various innervations, motor point entry sites, and even terminal intramuscular nerve endings, not to mention cortical representation…it is complex as you admit. We are just scratching the surface.

    However, let me pose a simple challenge. The next compliant person/patient that comes your way with reasonable motor control and minimal shoulder pain, palpate their lower trap and upper trap and see if you can make them turn on at different times. The same for the posterior and anterior deltoid. They function differently in my experience, no EMG crosstalk with the good old palpation. I do agree that it is relatively impossible to shut one down and turn one on completely. But let’s be clear that selective muscle BIAS has important clinical relevance whic seems to be dismissed ALMOST entirely in this post.

    Secondly, in surgeries, it is well known that the anterior deltoid is often lost while the middle and posterior deltoids are preserved secondary to anterior axillary nerve trauma, further evidence that they share seperate functions and have shared yet independent innervation, and that perhaps the anterior axillary nerve is more complicated than we initially thought.

    Another example. We can also use our digits independently (FDS, FDP, ED), although they share the same “muscle”. Same with the hamstring, fire the lateral and medial muscles seperately, we get tibial rotation in and out, much like the reigns of a horse moving the head back and forth.

    Would love your thoughts James.


  4. James Speck says


    Thanks for the thoughtful comment. It may well be shown in time that the trapezius has some level of functional differentiation. I’m very open to the idea that the CNS, in all its complexity, can fine tune the contraction of muscles within muscles. We can’t say presently we know this to be true for the trapezius, or even that there is good evidence to suggest this. Any talk about separate upper and lower trap activity is not well grounded.

    I place almost no value in palpation for detecting trapezius activity. There is no motion one can make that only activates the trapezius. Hence the reason trapezius manual muscle testing is impossible. The trapezius is a very thin muscle, paper thin in the cervical region, so we are never feeling only the trapezius contract under our fingers.

    The deltoid is an interesting case, and with palpation (as crude a test as it may be) I sense there may be different levels of activation in the anterior and posterior portions. Going beyond the 3 named sections, some studies suggest there may be 7 functional bundles within the deltoid, separated by intramuscular tendons and all (in theory) with the potential for selective activation by the CNS. One study even identified different levels of uptake in these sections with PET imaging. As far as I know, this is the closest anyone has come to establishing functional differentiation within what is commonly identified as a single muscle.

    I looked at several dissection studies when researching this post and none mentioned finding intramuscular septa in the trapezius. This is part of the reason why I’m skeptical of separately functioning upper/middle/lower portions.

    The loss of anterior delt function following surgery can have several causes (disruption of the attachment, cut intramuscular branches) so this may not in itself represent evidence of selective activation. But I agree the complexity of nerve function is beyond our current level of understanding.

    In regards to muscles that control the fingers, I’m not sure to what extent the extrinsic muscles (FDS, FDP, ED) can move a finger in isolation. Try to flex your pinky without the ring finger coming along for the ride. Intricate volitional finger motion relies on the interplay of agonist, antagonist, and synergist action of the extrinsic and intrinsic hand muscles. But there may be as you suggest separate functioning bellies within these muscles, with unique nerve suppliers, deserving of sub-classification. The hamstrings we identify as distinct muscles with clear anatomical separation. It’s a very good question to what degree each can be activated independently of the others, but this may not be relevant when discussing the trapezius which seems to be one continuous muscle in humans.

  5. Kris, PT says

    Thanks for the reply James, all great things to noodle on and I appreciate you bringing this topic up.

    Perhaps I shouldn’t trust my hands, but I have been trained and continue to fine tune my palpation skills and find them as important as my eyes and sometimes it’s helpful to let go of the studies and treat what you see and feel in front of you. But as soon as the patient leaves, go back to the research!

    I think the trapezius is larger than you are implying. The quadriceps tendon is also very small relative to the huge muscle that attaches to it. The mid-belly of my trapezius is 1″ thick easily. How thick it is in the neck does not give extensive insight into it’s function does it? I looked for an article on this but got tired, but I don’t think paper thin is remotely close to accurate (standard printer paper is .1mm thick or less). But on to the point. If you palpate the upper trapezius in the mid-substance and lower trapezius in the mid-substance you should not have a trouble discerning the two with practice.

  6. Felix says

    Is obvious that you did’t train with weights o whatever muscle training device. If you did for enough long you would learn -beyond nerve supply issue- that muscle fibers ALWAYS opperate like “elastic ropes” contracting gradually as the resistance is aligning to the fiber itself. When you are doing upper barbell/dumbbell press, you are targeting upper chest because those fibers are MORE aligned with the resistance vector (gravity) than the middle or low fibers. One more time Nature shows that FUNCTION CREATES ORGAN. Is too difficult for your understand that?

  7. alex says

    there’s only side of the equation in the article. so it’s wrong. muscles are smarter than just being triggered by nerve signals. i dont know exact biological mechanism but i’m a practicing bodybuilder and i know that muscles react to tension. tension depends on angles and leverages. so a muscle first recruits fibers in places of most tension. so it’s definitely possible to work on different fibers (not on fiber parts though, like lower, upper biceps). i can fell the upper chest and definite DOMS in upper chest fibers. so other side of the equation is tension receptors (or whatever that functions like that)

Leave a Reply

Your email address will not be published. Required fields are marked *