Why the dumbbell row should not be in your routine (and what you should do instead)

Exercise selection is an important consideration when designing a program, but remains poorly understood. Many trainees unfortunately rely on the advice of authoritative figures who have no right being as such, which is why many are still uninformed on the subject of selecting exercises that effectively target the musculature they want to grow.

Since the dumbbell row has remained a staple in the pro bodybuilding community, many still consider it a great exercise for growing the shoulder extensors (latissimus dorsi, posterior deltoid, teres major, and teres minor), your biceps, and the middle and lower trapezius. As you will see, however, this is not the case.

In this article, I’m discussing the variation of the dumbbell row shown below (though biomechanically similar narrow grip rowing exercises suffer comparable issues). The shoulder is extended parallel to your torso while using a bench for support.

Image result for dumbbell row

The dumbbell row: In the bottom position (left picture), the scapula is protracted with the arm perpendicular to the ground, and the elbow fully extended. At the top of the movement (right picture), the scapula is fully retracted, the shoulder is extended to 90°, and the elbow is flexed 90°.

Remember, its about progressive overload

Since progressive overload is key for muscle growth, its imperative to select exercises that lend themselves well to this fundamental principle. After all, if you aren’t adding resistance or adding reps to an exercise over time, I can assure you, you aren’t gaining an ounce of muscle.

Like every new exercise inserted into your routine, the dumbbell row probably “feels” like it’s working well at first. You get a little sore from the exercise, you feel a good pump from it, and you can even make relatively easy jumps in weight over the first few sessions, all is good. Or so you think. However, you shouldn’t let your subjective feelings get in the way. Your ability to add weight and reps to a new exercise is initially much more neurological rather than morphological. Every exercise is easy to make progress on at first, don’t let this initial honeymoon fool you.

Every exercise also becomes more and more challenging as you progress. Still, the dumbbell row becomes exponentially more challenging, to the point where many hit a plateau or develop major technical errors and cheat the movement. The main reason is due to the mismatch between the shoulder extensor strength curve and the shoulder extension resistance curve. These curves are graphical representations of how the resistance torque and muscular torque production vary over the range of motion (ROM) of an exercise.

To understand how each curve is generated, you need to have a basic understanding of torque.

Basic principles of torque

Torque is a force that causes an object to rotate. Since the execution of virtually every exercise involves rotating your skeletal components about their articulation, torque is the name of the game.

Two variables determine the torque acting on an object:

  • The Force (F) acting on the object. In the case of resistance that generates torque during an exercise, the force is often the weight of the loaded bar, the dumbbell, your body, etc.
  • The Moment arm (MA). The MA is the perpendicular distance from which a force is acting to the point of rotation (the fulcrum).

The torque (T) acting on an object is the product of the force and the moment arm:

\text{ T } =  F * M.A.

To execute the concentric phase of an exercise, your muscles need to generate an opposing torque that is greater than (or equal to, once the weight is moving) the torque generated by the resistance.

If weight were the only variable that determined resistance, every exercise would have a constant resistance curve because the weight doesn’t change during an exercise (unless you’re doing accommodating resistance work with bands, chains, or something else). The moment arm between the weight and axis of rotation (your joint), however, is generally not constant throughout the ROM for many exercises. As a result, many exercises have a resistance curve that is not constant.

Similarly, the internal moment arm a muscle uses to generate torque typically changes over the ROM of an exercise.

Dumbbell row resistance curve

When you execute the dumbbell row, the shoulder extensors are subjected to a descending resistance curve due to the circular arc the dumbbell travels, along with the constant vertical force from gravity.

A descending resistance curve means resistance torque is lowest when the muscles are in their lengthened position but are then subjected to an ever-increasing resistance torque throughout the concentric phase of the exercise until resistance torque reaches a maximum at the end of the concentric phase.

Here is the resistance curve.

(Reminder: The ROM for an exercise is measured in degrees of joint rotation, not the linear displacement of the dumbbell, barbell, machine weight stack, etc.)

Lets now examine how the strength curve matches up with the resistance curve.

Force production of the shoulder extensors

Within a muscle, force production is the sum of active and passive forces. Muscular force production is typically not constant over a given biomechanical ROM.

Active force is determined by how lengthened or shortened the individual contractile units (sarcomeres) are within the muscle. There is a maximum force-producing sweet spot in which the sarcomeres are at an ideal length.

A decent way to visualize when your muscles are near the ideal length and capable of producing peak active force is when your body is in an anatomically neutral position (think military posture with your elbows, knees, and hips slightly flexed). As you deviate from the ideal force-producing length by contracting or elongating a muscle, the ability of the muscle to produce active force diminishes.

However, depending on the muscle, the amount sarcomeres are lengthened and contracted throughout a given biomechanical ROM varies.

Gerling et al. investigated the working sarcomere length for the lats during shoulder extension, and Interestingly, sarcomeres within the lats appear not to change in length much throughout the dumbbell row ROM. The study also tells us the lats operate primarily within the ideal length region during the dumbbell row. Therefore, the lats can produce near peak active force throughout the entire ROM for the exercise.

Unfortunately, there isn’t research examining operating sarcomere lengths for the posterior deltoid, teres minor, and teres major during shoulder extension. However, they are also likely able to produce high levels of active force, especially near the top of the dumbbell row, since the top position corresponds to anatomical neutral.

As for passive forces, they don’t come into play much during the dumbbell row. The elastic structures of the muscle generate passive force during the eccentric phase of an exercise once sarcomeres are lengthed beyond the ideal length. Based on the Gerling et al. study, sarcomeres within the lats won’t extend much beyond the ideal length during the exercise. It’s also likely the posterior delt, teres minor, and teres major won’t either, and there is minimal resistance at the bottom of the dumbbell row anyway.

Lack of passive force production is a disadvantage for the dumbbell row and other exercises that possess a descending resistance curve. When you load a muscle in its stretched position, you achieve more overall muscle growth, and you also promote muscle hypertrophy in series (lengthening of the muscle belly).

While we’re on the topic, the limited ROM of the exercise is also a shortcoming for muscle growth and strength development. The shoulder extensors are capable of producing torque over much greater ROM (think ROM for chin-ups). Full range of motion training leads to more muscle growth and better strength development [2,3].

Internal moment arms & torque production of the shoulder extensors

Since the shoulder extensors can produce a high amount of active force throughout the exercise, you would think they can handle the demands of the exercise. Remember, though, muscular torque production is what ultimately determines the strength curve, so we need to understand how the internal moment arms of the shoulder extensors change over the ROM.

A review by Ackland et al. showed the internal moment arm for the lats, teres minor, and teres major decrease over the last half of the ROM (45°). The thoracic region of the lats suffer about a two-fold reduction in moment arm over the last half of the exercise, which is likely the region of the lats that contributes most to shoulder extension. The internal moment arm of the posterior deltoid does increase over the last half of the exercise, however.

The pectoralis major is also a shoulder extension synergist during the dumbbell row. Leverage of the pectoralis major also decreases throughout the concentric phase of the dumbbell row (and even becomes a shoulder flexor near the top) [2].

What does this tell us? Overall, the increase in leverage of the posterior delt likely doesn’t offset the loss in leverage of the lats, teres major, teres minor, and pectoralis major.

Given the data we have, the strength curve probably peaks around the halfway point of the exercise, then decreases over the final half of the exercise, creating a mismatch between the strength curve and resistance curve once the weight is high enough.

Ideally, you want the resistance curve to match your strength curve. When you fail an exercise due to a mismatch in strength vs. resistance, other portions of the exercise remain understimulated.

And no, this does not mean its ok to perform the exercise with only the first 45° of ROM. You’d be performing a very limited ROM exercise (this is why we don’t do half squats) that still has a descending resistance curve.

Moreover, as you add more weight to the dumbbell row over time, the resistance curve becomes steeper. Consequently, the mismatch in strength vs. resistance likely becomes more pronounced over time, especially since strength gains diminish over time. This is why many struggle to make continuous progress without reducing ROM or using egregious amounts of momentum by twisting their torso or jerking their body in an attempt to complete the lift. Sound familiar?

Is using momentum a viable option?

Using momentum during the concentric phase allows you to use more weight than you otherwise could, allowing you to emphasize the eccentric phase of the exercise because your muscles are stronger when they are lengthening. This training technique is called eccentric overload and has been shown to benefit muscle growth and strength development.

However, it’s not practical to implement eccentric overload with the dumbbell row.

It’s easy to get carried away with adding too much momentum assist via hip extension and axial twisting over time, along with allowing the dumbbell to lower at a faster rate (did you actually get stronger, or did you just get better at throwing the weight up while allowing it to drop faster?).

It’s also likely you’re not taking full advantage of the eccentric overload technique due to the resistance curve of the dumbbell row. Eccentric overload is best suited for exercises that allow for full ROM and requires the muscles involved to produce a high amount of force throughout the entire eccentric phase of the movement, including when the muscle is in its stretched position. You’re not able to accomplish this with the dumbbell row.

Many also consider the dumbbell row a good biceps exercise. But is it?

The dumbbell row is hardly a biceps exercise if you think about it. There is virtually zero resistance the biceps need to overcome throughout the ROM due to the constant vertical forearm. The biceps are also taken through a limited ROM (~90°) during the exercise.

A study by Mannarino et al. showed that dumbbell biceps curls resulted in more muscle growth than supinated dumbbell rows. You may think this doesn’t necessarily prove dumbbell rows are an inadequate biceps exercise, but we also have this study by Gentil et al. that showed wide grip pulldowns resulted in the same biceps growth as barbell biceps curls. Chin-ups and pull-ups are different from the dumbbell row; they create elbow extension torque the biceps need to overcome to flex the elbow.

Greater muscle growth from doing the biceps curl vs. dumbbell row isn’t due to the fact you’re performing a biceps isolation exercise, its because the dumbbell row doesn’t train the biceps well compared to other compound pulling movements like pull-ups, chin-ups, and pulldowns.

Trapezius development

Lastly, some argue the dumbbell row is still a good exercise for the middle and lower traps because you’re protracting and retracting the scapula through its entire ROM. The resistance curve for the traps during the dumbbell row is constant, which isn’t as bad as a descending resistance curve. However, you’re only as strong as your weakest link during an exercise. Once you hit a plateau from your inability to extend your shoulder, you cut off progress on the traps as well.

What you should do instead of the dumbbell row

Here are some exercises that do a better job training some or all of the muscles used during the dumbbell row:

  • Chin-ups/Pull-ups: Classics that have stood the test of time. They offer better resistance curves, more ROM for the lats, posterior delt, and synergists, and are much better for the biceps.
  • Wide-grip machine pulldowns (video): An exercise well suited for eccentric overload. The fixed grip width produces a constant shoulder adduction resistance curve, loading the shoulder adductors throughout the entire eccentric phase. You achieve near full ROM for the lats, posterior delts, and shoulder adduction synergists.
  • Reverse Bayesian fly (video): A good exercise for the posterior delt and the middle and lower trapezius, but not the lats. The resistance curve is bell-shaped; the muscles produce tension over full ROM for the movement with resistance peaking at your strongest point. You can play with the cable height and distance you stand from the cable tower to alter the resistance curve to find what’s best for you.
  • Lat prayer (video): A lat dominant exercise that offers full ROM and a bell-shaped resistance curve. The lats are also loaded in the stretched position.


Compound exercises are an important staple of a good training program. However, not all are equal in their ability to stimulate muscle growth for the muscles involved, nor equal in terms of their ability to allow for continuous progress.

The dumbbell row provides an inferior stimulus for muscle growth and is not suited well for progressive overload:

  • The shoulder extensors are taken through a limited ROM.
  • The shoulder extensors are subjected to a descending resistance curve; the passive structures within the muscles are not loaded, which limits muscle growth.
  • Progressive overload becomes nearly impossible to implement as a consequence of the poor stimulus for muscle growth and strength development, along with an ever-increasing mismatch between the strength and resistance curve over the last half of the exercise.
  • The biceps are poorly stimulated during the exercise.

Of course, there’s always the classic objection “but person X at my gym has a massive upper back, lats, and biceps and swears by the dumbbell row.” This is likely a case of correlation does not imply causation, my friend. If you or someone else are doing other exercises in your routine that targets those muscle groups, how do you know with certainty the dumbbell row is contributing to their development? You don’t.

Also, consider this, biomechanical differences and responsiveness to training play a role in one’s ability to progress on an exercise, regardless of how inferior the exercise may be.

If someone is an excellent responder to training, such that even the lousy stimulus the dumbbell row offers still induces enough strength development, along with having a more favorable biomechanical advantage (read: this probably isn’t you), they’ll make more progress versus someone who is not as good of a responder to training in general. However, this doesn’t mean a genetic freak wouldn’t be better off doing other exercises that offer an overall better stimulus for muscle growth.

Leave a Comment

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