Bench Press Range of Motion: An Exception to the Principle of Specificity?

Much of what we know about range of motion specificity comes from single-joint studies and squat studies. When we branch out to the bench press, things get more complicated.

bench press range of motion specificity

This article is a review and breakdown of a recent study. The study reviewed is Bench Press at Full Range of Motion Produces Greater Neuromuscular Adaptations than Partial Executions After Prolonged Resistance Training by Martínez-Cava et al (2019)

Key Points

  • Subjects trained for 10 weeks, doing either full bench press reps or one of two partial ranges of motion (⅓ reps or ⅔ reps). They tested strength and velocity at all three ranges of motion pre- and post-training.
  • Unexpectedly, the full range of motion group tended to improve the most in all measures at all ranges of motion, not just the full range of motion measures. The ⅓ range of motion group tended to improve the least in all measures, even for the ⅓ range of motion tests.
  • While the principle of specificity has a tremendous amount of support, we need to remember that it’s a principle, not an iron-clad law of the universe. In the interpretation section, I’ll discuss when it may or may not apply.

One of the first things you learn about when you start consuming strength training content is the principle of specificity. The principle of specificity has wide-reaching implications, but one of the well-known applications is range of motion specificity: you gain the most strength in the range of motion you train for. In other words, if I want to improve my deep squat, I’d want to do deep squats, but if I want to improve my half squat, I’d be better off doing half squats.

However, we need to keep in mind that the principle of specificity is more of a strong heuristic rather than an iron-clad law of the universe. Sometimes, it doesn’t apply. And when it doesn’t, we can learn something by thinking through the factors that may be able to “override” such an important principle.

In the present study (1), three groups of subjects trained the bench press through either a full range of motion, a ⅔ range of motion, or a ⅓ range of motion, with strength and velocity testing for all three ranges of motion pre- and post-training. The full range of motion group improved the most for tests through all ranges of motion, while the ⅓ range of motion group got the worst results, including on the tests in the range of motion they were actually training. The interpretation section will dig into factors that may explain why the results of this study run counter to what we’d expect, given the principle of specificity.

This article is from a previous issue of Monthly Applications in Strength Sport (MASS), our monthly research review with Greg Nuckols, Eric Trexler, Eric Helms and Mike Zourdos. Every month, we analyze 10 of the most important studies for strength and physique athletes and coaches, then write about the practical applications in concise, jargon-free research reviews (like this one!). 

If you're interested in staying up to date with the latest strength and nutrition research, you can find the MASS subscription plan that works for you and subscribe here.

Purpose and Hypotheses


The purpose of the study was to investigate the effects of bench press range of motion on strength and velocity adaptations.


No hypotheses were given.

Subjects and Methods


49 young men who had been benching 2-4 times per week for at least 6 months completed the study. More details about the subjects can be seen in Table 1.

Experimental Design

Before any performance testing, the subjects underwent nine familiarization sessions: three sessions of ⅓ bench reps, three sessions of ⅔ bench press, and three sessions of full bench reps. After familiarization was completed, the subjects underwent three testing sessions. In each session, they completed a load-velocity profile and a 1RM test with one bench press range of motion (i.e. ⅓ reps, ⅔ reps, or full reps). The order of the testing sessions was randomized for each subject, and the same order was repeated for post-testing. The load-velocity testing started at 20kg, with loads increased by 10kg per set until mean propulsive velocity (2) fell below 0.5 m/s, after which time loads increased by 2.5-5 kg per set until a 1RM was reached. Subjects performed 3 reps per set with light loads (<50% 1RM), 2 reps per set with moderate loads (50-80% 1RM), and 1 rep per set with heavy loads (>80% 1RM).

After the initial testing sessions were completed, subjects were assigned to one of four groups in a counterbalanced fashion based on pre-training bench press strength. One group trained doing ⅓ reps, one group trained doing ⅔ reps, one group trained doing full reps, and a control group didn’t train at all. The three experimental groups trained twice per week for 10 weeks, using a linearly periodized training program. Loads were selected and adjusted using velocity targets that were intended to correspond with the target intensity for the day. Details of the training program can be seen in Table 2. After 10 weeks of training, the performance tests were repeated.

The researchers did a good job of standardizing as many aspects of the study as possible. The subjects trained using a Smith machine (this probably wasn’t necessary, but it does make velocity data a little more accurate since all of the movement is completely vertical, though this accuracy comes at the cost of a bit of ecological validity), grip width was standardized (5-7 cm outside of shoulder width), and safety bars were used to ensure that range of motion was appropriate and consistent for each rep. The subjects lowered the bar to safety pins and paused for two seconds before pressing each rep, making the demarcation between the eccentric and concentric crystal clear, thus theoretically improving velocity measurements and ensuring a consistent range of motion. Furthermore, subjects were instructed to maintain a velocity of 0.45-0.65m/s for their eccentrics (with the aid of visual and audio feedback from the velocity device used in the study), and to press each concentric as explosively as possible.

Prefer to listen? Check out the audio roundtable

Every article we review in MASS has an accompanying audio roundtable just like this one. All of the MASS reviewers (Greg Nuckols, Eric Trexler, Eric Helms, and Mike Zourdos) get together to discuss the findings and applications in practical, easy-to-understand terms. Subscribe to MASS here


Surprisingly, the group doing full reps tended to improve the most in all measures, and the ⅓ reps group tended to improve the least. According to the principle of specificity, one would have expected that the full reps group would improve the most at full reps, the ⅔ reps group would improve the most at ⅔ reps, and the ⅓ reps group would improve the most at ⅓ reps. However, that was not the case. Figure 1 tells the story. Note that not all differences between groups were statistically significant (and there are 72 potential pairwise comparisons; it’s not worth going through all of them one by one), but the overall pattern is crystal clear.


Upon reading the title of this article, I thought it was going to be a straightforward article that I could review in my sleep. “There’s this thing called the ‘principle of specificity,’ here are all of the studies backing it, and the present study adds one more to the pile.” In and out, easy peasy.

However, this study does not fit that narrative. Training through a full range of motion was best for improving performance through a full range of motion, but training through a partial range of motion wasn’t best for improving performance through a partial range of motion. In fact, the ⅓ range of motion group had the worst gains though the ⅓ range of motion in three of the four measures. So, I can’t just type out my stock “principle of specificity” spiel and call it a day. However, while that makes this article more difficult to interpret, it also makes it much more fun.

Before I get ahead of myself, let’s just recap the results: a longer range of motion seems to be the way to go for … everything, at least in this study. When I thought about it for a moment, I realized that the key studies people generally cite to support the concept of range of motion specificity use the squat (3) or single joint exercises (4), rather than the bench press.

What about the bench press specifically? There are three other relevant studies, but only one of them tested strength through multiple ranges of motion. Clark et al compared the effects of benching through a full ROM against benching through a variable ROM (a combination of full reps, ¼ reps, ½ reps, and ¾ reps; 5). Neither group experienced significant increases in force output through a full range of motion (which calls into question the usefulness of the training program used in the study), but the variable ROM group had a larger increase in force output through a ½ rep ROM. Two studies by Massey et al were nearly identical, except that one of the studies used male participants (6) while the other used female participants (7). Both studies had three groups: one trained through a full range of motion, one group trained through a partial range of motion (not lowering the bar below the sticking point), and one group did half of their sets through a full ROM and half though a partial ROM. In both studies, full-ROM bench press strength was the only outcome measure. In the study on males, the full and partial ROM groups had similar increases in strength, while the group doing both full and partial reps tended to gain a bit less strength. In the study on females, on the other hand, the full ROM group had the largest increase in 1RM strength, while the partial ROM and mixed ROM groups had slightly smaller increases. Putting all four of these studies together, it becomes clear that doing full ROM training is important for building strength through a full ROM on bench press (benching through a full ROM built strength through a full ROM or through the bottom of the ROM as well as or better than all other conditions in all studies). However, ROM specificity isn’t as clearly supported. In the present study (1), benching through a full ROM built more strength through a partial ROM than benching through a partial ROM did. In the Clark study, while partials built more strength through the top part of the ROM, benching through a full ROM failed to increase strength off the chest to a greater degree than doing partials. In the Massey study on males, benching through a full ROM and doing partials proved equally effective for building strength through a full ROM. And finally, in the Massey study on females, the group using mixed ROMs (and thus still doing some training through a full ROM) failed to increase full ROM strength more than the group only doing partials. All in all, it’s a murky picture.

So, what might explain these results? If the principle of specificity is so well-supported, why does ROM specificity in the bench press look so iffy?

My first thought is that the magnitude of strength fluctuations throughout a full range of motion probably has an impact. In other words, are you 20% stronger at the top of a rep than at the bottom of a rep, or are you 100% stronger? In the case of squats, it’s not terribly uncommon to be able to lift WAY more for partials than for full reps. Speaking from experience, I’ve done quarter squats with 1000+lb when my 1RM through a full ROM was closer to ~650, and I’m sure that disparity would be even larger if I trained quarter squats with as much focus as I put into full ROM squats. With bench, on the other hand, there’s maybe a 15% difference between my full-ROM bench press and the heaviest weight I could use for high pin presses or board presses. The exact ratios may differ for you, but I’d be surprised if that same principle didn’t apply to almost everyone reading this: there’s a bigger gap between partial-ROM strength and full-ROM strength in the squat than in the bench press. By extension, one would then assume that full-ROM bench press would have a larger effect on partial-ROM bench press strength than full-ROM squatting would have on partial-ROM squat strength. If an appropriate load for full-ROM bench press is 200lb and an appropriate weight for partials is 230, doing reps at 200 is still probably heavy enough to do something for partial-ROM strength. However if an appropriate load for full-ROM squats is 400lb and an appropriate load for half squats is 650, doing full-ROM reps with 400 probably isn’t doing much to improve partial-ROM strength directly (beyond simply building more muscle mass). The opposite principle may also be true – you may get better carryover from partial-ROM training to full-ROM strength when the strength curve of a movement is flatter. Half reps on bench press feel more similar to full-ROM bench press than half reps on squat feel when compared to full-ROM squats. If you’ve ever converted an athlete from partial-ROM training to full-ROM training, you’ve probably seen this firsthand – their first session benching through a full ROM is a little humbling because they need to take some weight off the bar, but they still perform reasonably well, and the weights they can handle are at least somewhat comparable to the loads they were using before for partials, unless they were previously using a very partial ROM. With squats, on the other hand, shifting from half squats to full squats often requires a complete rebuild of their squatting mechanics, and necessitates slashing their training weights at least in half.

In the present study (1), another factor in play is how the subjects actually performed their reps. To keep ranges of motion consistent, the subjects benched to pins, allowing the bar to briefly rest on the pins between sets. This is a great way to ensure that the range of motion was consistent, but it possibly decreases ecological validity a bit. When you switch between the eccentric and concentric portions of a lift, if there’s not a physical impediment to help you decelerate the bar (i.e. the floor when deadlifting), you get a rather large spike in force output when transitioning between lowering the bar and explosively lifting the bar. Lowering the bar to pins negates that spike in force output. That does mirror the way that many people would do partial bench press reps (i.e. pin press or board press), but it may not capture all of the ways that someone could apply partial range of motion training when bench pressing (namely, simply reversing each rep yourself without touching your chest). It’s plausible that partial ROM exercises that require the athlete to actively decelerate and reverse the load build more partial ROM strength than exercises, like pin presses, that allow another physical object to decelerate the load.

Finally, it’s important to think about these results conceptually, rather than simply accepting one study as the final word on the topic. We have good research indicating that longer ranges of motion tend to lead to more muscle growth (38). Since the subjects weren’t very well-trained, they likely still had plenty of room to grow more muscle. Hypertrophy wasn’t assessed in this study, but I don’t think we’d be unjustified to assume that that full range of motion group likely experienced the most muscle growth. That could be enough to explain the superior gains in performance, even through partial ranges of motion. Reasoning by analogy, two studies on squatting and jump performance immediately come to mind. In one study on untrained lifters (3), deep squats led to greater improvements in jump height than half squats (the positioning of half squats more closely mirrors jumping mechanics than the positioning of full squats). In another study on high level athletes (9), half squats led to larger improvements in jump height than full squats. In my opinion, the most likely explanation is that for untrained or semi-trained athletes, more hypertrophy can occur, and that muscular development can lead to robust performance improvements. For highly trained lifters, on the other hand, much less hypertrophy can occur, so optimizing for movement specificity rather than hypertrophy leads to larger performance improvements. Thus, if this study was repeated on trained lifters, I would expect that specificity would apply to a greater degree, with the full ROM group having the largest strength gains through a full ROM, the ⅔ ROM group having the largest strength gains through a ⅔ ROM, and the ⅓ ROM group having the largest strength gains through a ⅓ ROM.

Finally, I’d just like to touch on something the authors of the present study mention in their discussion (1). They propose that powerlifters are actually training the bench press through a partial range of motion, due to arching, taking a wide grip, and specifically aiming to minimize range of motion, and that they may be able to gain more strength if they did more of their training through a purposefully longer range of motion. I think that’s an idea worth at least considering. Based on competition definition, a “full” range of motion is any range of motion that allows the bar to touch your chest and lock out, as long as your grip width doesn’t exceed 81cm. But is that REALLY a full range of motion biomechanically? I’d argue that it isn’t. For example, a close grip bench with a smaller arch involves more elbow flexion, and a greater combination of shoulder extension and horizontal abduction than a typical competition-style bench, so your prime movers clearly aren’t going through their full range of motion with a wide-grip, arched bench press. Is it possible that optimizing technique for short-term performance could actually limit long-term development, assuming you do most of your bench press training with a competition-style set-up? I certainly think it’s possible. For what it’s worth, that matches my experience (I always tend to make better bench progress when I’m doing a lot of cambered bar bench or close-grip bench with a smaller arch). It matches the anecdote of Mike MacDonald, who may be the most successful bench presser of all time; he simultaneously held the bench press records in four different weight classes at one point and swore by cambered bar bench press. More recently, Josh Bryant’s lifters have been very successful on the bench press, while primarily benching with pretty narrow grip widths (at least by powerlifting standards). Jeremy Hoornstra and Julius Maddox are his two most successful lifters, owning the all-time bench press records at 242, 275, and superheavyweight. At the very least, if your bench press is plateaued, I think it’s worth considering doing some of your weekly bench press training with a technique that allows for a longer range of motion.

Next Steps

I’d like to see more research looking at range of motion specificity in a wider array of exercises and in more advanced lifters. I’d also like to see a training study in powerlifters comparing a training program consisting solely of wide grip bench against a training program with pressing volume split evenly between wide-grip bench and close-grip bench.

Application and Takeaways

While the principle of specificity is a cornerstone of training theory, it’s important to remember that it’s a principle, not an iron-clad law. Specifically, range of motion specificity may not hold up quite as well in the bench press as in the squat. For long-term strength development, benching through a longer range of motion than your competition-style setup may be worth a shot if you plateau.

The latest research – interpreted and delivered every month

If you want to stay up-to-date on the research pertinent to strength and physique athletes and coaches, but you don't have the time or desire to develop the skill set to critically analyze research, you can sign up for Monthly Applications in Strength Sport (MASS), the monthly research review I put out every month, along with Dr. Eric Trexler, Dr. Eric Helms, and Dr. Mike Zourdos.

Each issue of MASS contains at least 10 pieces of content like this. Click here to learn more and join 3,500+ subscribers


  1. Martínez-Cava A, Hernández-Belmonte A, Courel-Ibáñez J, Morán-Navarro R, González-Badillo JJ, Pallarés JG. Bench Press at Full Range of Motion Produces Greater Neuromuscular Adaptations Than Partial Executions After Prolonged Resistance Training. J Strength Cond Res. 2019 Sep 26.
  2. Mean propulsive velocity is very similar to mean concentric velocity. The difference is that mean propulsive velocity trims off the end of each concentric when the bar decelerates prior to lockout before calculating average velocity, whereas mean concentric velocity is a measure of velocity for the entire concentric portion of each rep. At high loads, mean propulsive velocity and mean concentric velocity are similar. At low loads, mean propulsive velocity is a bit faster than mean concentric velocity since the bar needs more deceleration prior to lockout. For the purposes of this study, the differences don’t really matter, since both are valid and reliable measures and convey nearly identical information.
  3. Bloomquist K, Langberg H, Karlsen S, Madsgaard S, Boesen M, Raastad T. Effect of range of motion in heavy load squatting on muscle and tendon adaptations. Eur J Appl Physiol. 2013 Aug;113(8):2133-42.
  4. Valamatos MJ, Tavares F, Santos RM, Veloso AP, Mil-Homens P. Influence of full range of motion vs. equalized partial range of motion training on muscle architecture and mechanical properties. Eur J Appl Physiol. 2018 Sep;118(9):1969-1983.
  5. Clark RA, Humphries B, Hohmann E, Bryant AL. The influence of variable range of motion training on neuromuscular performance and control of external loads. J Strength Cond Res. 2011 Mar;25(3):704-11.
  6. Massey CD, Vincent J, Maneval M, Moore M, Johnson JT. An analysis of full range of motion vs. partial range of motion training in the development of strength in untrained men. J Strength Cond Res. 2004 Aug;18(3):518-21.
  7. Massey CD, Vincent J, Maneval M, Johnson JT. Influence of range of motion in resistance training in women: early phase adaptations. J Strength Cond Res. 2005 May;19(2):409-11.
  8. McMahon GE, Morse CI, Burden A, Winwood K, Onambélé GL. Impact of range of motion during ecologically valid resistance training protocols on muscle size, subcutaneous fat, and strength. J Strength Cond Res. 2014 Jan;28(1):245-55.
  9. Rhea, M., Kenn, J., Peterson, M., et al. Joint-Angle Specific Strength Adaptations Influence Improvements in Power in Highly Trained AthletesHuman Movement, 2016, 17(1), pp. 43-49.


Scroll to Top