Over the years, whenever I’ve been in the weight room with young male tennis players, the following question always seemed to come up - “when are we going to bench?”. No surprise here as most teen boys are eager to work on their ‘pecs’ (and in case you’re wondering, it’s not to improve tennis performance).

But is there a place for the bench press in the training programs of tennis players? There are many that believe it’s completely useless; some even take it to the point where programming push-ups for players is taboo (FYI, doing push-ups for tennis isn’t bad; doing poor push-ups, with incorrect technique, at the wrong times, and in excess, is bad...but more on that rant in another post). Others do bench 3x a week, trying to get that ‘pump’ feeling.

Who’s right? As with most things, there’s no fast and easy answer. This article will take a closer look, investigating some of the research on the topic (and why we would even consider bench pressing in the first place).

Agassi (and Other Pros)

If you’ve ever read Andre Agassi’s book, Open, you would have come across a section on Andre’s training regime. He claimed to have lifted over 300lbs on the bench press (which is over 130kg) - and some sources have suggested that during his resurgent years, his body mass was between 165lbs and 180lbs...which means that he was benching close to 2 times his bodyweight (not bad Andre!).

Whether that’s true or not, isn’t the point. The point is that he had a strong and powerful upper-body. And he’s not the only one. Others have released videos and photos of themselves benching (check out the video of Tsonga below). It’s not just male players either. I’ve personally worked with female players that have asked to perform some benching during preparation periods.

As you can see, Agassi was quite lean but did have a ‘well-developed’ chest.

As you can see, Agassi was quite lean but did have a ‘well-developed’ chest.

Let’s get one thing straight though before we move on and look at the research on this topic. It’s important to note that Andre didn’t do sets of 10 reps at 300+ lbs (or any other weight for that matter). In fact, Gil Reyes reported in an old article that the 8-time Grand Slam champ was doing as many as 7 sets when benching and in rep ranges that were quite low (between 2 and 5). He even said they would avoid loads where Andre ‘struggled’ on the last rep or two.

Why is this important? Well, if you’ve followed any of my previous work, you’ll likely know that these rep and set schemes primarily target the max strength end of the resistance training spectrum. In other words, you can get quite strong doing things this way, without any significant changes in body mass. Recall that this type of training principally targets neural aspects. So it’s not surprising that Andre could maintain his lean physique, but still gain appreciable levels of strength.

Working to failure - i.e. performing sets until you can barely lift the last rep (that burning sensation) - causes a significant amount of damage at the cellular level. Not only does this contribute to an increase in fiber size (when performed long-term of course) but you can get more systemic fatigue with this type of training. Neither are preferred when it comes to training & recovery for tennis.

But just because Andre benched, doesn’t mean we should copy his program blindly. Was there a reason behind Gil Reyes’ prescription? Is upper-body strength via the bench press a worthwhile pursuit?

The Bench Press and Tennis Research

There isn’t a whole lot of research when it comes to the bench press and tennis performance (to my knowledge, none that directly compares bench pressing to some sort of performance factor).  One study (Terraza-Rebollo et al 2017), did find that after 8 weeks of training, high-performance junior players that performed a 3 days/week strength training program, saw significant increases in serve speed (an overall 6km/h increase). There were 2 other groups - one that performed a MB throwing routine 3 times/week and a control group that only did the scheduled tennis sessions 3 times per week (all 3 groups did the same tennis training). Neither of those groups improved serve or groundstroke speeds.

While the strength group did incorporate the horizontal barbell bench press in 2 of the 3 sessions, there were another 8 exercises included on each of the days, so it’s tough to say whether the increase in serve speed was due to benching, some other exercise or a combination of exercises.

Note - This is why when we program our lifting sessions for tennis players, we usually only have 1 or 2 ‘main lifts’ during 1 session. And in general, there aren’t more than 3-5 (maybe 6) exercises that we go through (depending on time of year and player needs). If your strength & power sessions have 9 or more exercises, how can you accurately assess whether something is working or not?

Below is an example of ATP player, Chris Heyman’s strength & power program during his prep period for 2019. For various reasons, I didn’t get him to do a traditional BB bench press (he did DB incline press instead). Even still, look at the sets/reps & take not of the tempos (T-U - tempo up; T-D - tempo down; X - explosive, C - controlled, S - slow).

Screen Shot 2019-06-02 at 12.10.31 PM.png

In another study (Kraemer et al 2003), collegiate female players underwent a 9-month, non-linear periodized resistance training program. In terms of ball velocities, they all (forehand, backhand and serve) increased significantly at the 4, 6 and 9 month marks. While their 1-RM bench press also increased at each interval, Kraemer had this to say in terms of the correlations of upper-body force and ball velocities:

“It is possible that resistance training increased upper-body force production capabilities sufficiently to enable a better transfer into power and high velocity force production necessary during the different tennis strokes. The increases in maximal strength in the shoulder musculature were important as it was previously reported that strength and joint laxity were the only significant contributors to ball velocities of different tennis strokes with the 1-RM shoulder press being more highly correlated (R 0.69) to three ball velocities than the bench press (R 0.30) (20). This indicates that the strength of the deltoids appears more important to the development of higher ball velocities than the pectoralis muscles of the chest.”

In my opinion, the authors are jumping to conclusions here that aren’t adequately backed by facts. In this particular study, no correlations were made between ball velocities and various strength measures, including the shoulder press (not to say that this exercise isn’t a contributor, but we just don’t know for certain if it was the most important one). Also, the serve is complex, it requires contributions from the entire body, at different times during the stroke. Could the ball velocities have improved because of the entire lifting program (which included heavy squats, single-leg work and more)?

Lastly, the study that they refer to in the quote above (where shoulder press was correlated but bench press wasn’t), dealt with collegiate female players that had very little lifting experience. Their 1RMs were, on average, 34kg for the bench and 24kg for the shoulder press. Those are numbers that I get 14 year olds to achieve within several weeks of training. So, the results of that study cannot be generalized as those players were extremely weak to begin with.

Research on Bench Press in Other Overhead Sports

Are there similarities between throwing and serving? Are there contributions from the pectorals to hitting? These are things we should know first if we’re going to look at whether research in throwing & benching is at all relevant for tennis (and whether or not we just begin to implement this exercise).

Well, Wagner et al (2012) did compare the tennis serve with the volleyball spike and the handball throw. What that study revealed was that, although there are unique qualities that define each of the aforementioned movements, they do all have a ‘general motor pattern’. In other words, they have the same distal to proximal execution sequence - whereby the movements of the pelvis, trunk and arm (and the angles with which these body segments work through) are very similar. That’s good to know.

On top of that, when looking at the acceleration phase of various overhead throwing and striking sports, what we often see are EMG (electromyography) readings of the pectoralis major that are much higher than any other muscle group. Below is a table that showcases this:

Adapted from Escamilla and Andrews 2009;    MVIC - maximal voluntary isometric contraction

Adapted from Escamilla and Andrews 2009; MVIC - maximal voluntary isometric contraction


1) these are all average values and in certain cases, sample sizes were small (8 in the tennis study, for instance); because of this, caution should be taken when using inferences.

2) Over 100% MVIC can occur in certain cases (for instance, when a subject is unable to contract maximally during an isometric test - usually this occurs because they are untrained).

3) I have not come across any studies that look at EMG activity during baseball batting, but I would guess that it would be similar to a tennis forehand and/or golf swing.

Ok, so now that we see that in fact there are similarities between throwing and serving and that the pecs play a role (to some capacity) in striking, hitting and throwing, let’s look at a couple more studies to see what we can find.

Before we continue, in case I haven’t said this yet (I probably have in previous posts), the reason we’re even talking about any of this is that we’re looking at increases in upper-body strength & power - and whether those increases play a role in hitting power. And to increase power, we either have to apply more force to the implement (racquet in our case) or increase the speed of muscle contraction. Both of these, in theory, can be manipulated through the use of weight room pressing activities.

In a 2014 study (Loturco et al), the authors noticed that elite karate athletes had a strong correlation between both the bench press and the bench throw (an exercise we’ll tackle in another post) AND punching acceleration. Believe it or not, punching in karate, boxing etc, have similar dynamics with hitting a forehand in tennis - just a side note.

In handball, throwing velocity is a key determinant to success and is why Hoff and Almasbakk (1995) conducted a training study with elite female handball players to see whether benching could improve throwing speed. They bench pressed 3 times a week for 9 weeks, progressively increasing load during that time period. It’s important to note (actually really important), that during the training period, the players continued handball training - which obviously included handball throwing.

At the 10 week mark, when the subjects were re-tested, the bench press group AND the control group (no weight training), both increased running and standing throwing velocities. But there was a significant difference here - the bench press group, because they got stronger, improved their throwing velocities to a higher degree (13kg increase in strength coupled with a 3.5m/s [standing] and 3.9m/s [running] increase in throwing speed).

In another study (Marques et al 2007), elite male handball players showed a positive and high correlation between throwing velocity and both 1-RM bench press and bench press bar velocities (we’ll look at why bar velocity is important in another post).

A couple notes from these studies. First, in both cases (training and testing), athletes are instructed to move the loads with maximal intent. In other words, the aim is not just to get the bar up to the finish point, it’s to get it there as quickly as one can.

Furthermore, in the training study, rep schemes were low (no higher than 6) and the loads were increased by 2.5kg if an athlete could do more than 6 reps comfortably. Let’s take a look at why all this is important.

Practical Considerations

I’ve said it before and I’ll say it again, when it comes to lifting weights for performance purposes (and for most sports, speed reasons), it’s not important how fast the actual movement is, but that the intent to move the load is maximal. Internally (i.e. muscle fiber level), if we try to move the load as fast as possible throughout the entire concentric portion of the lift, the contraction speeds are actually fast - and so are the recruitment requirements.

Just listen to the video of Tsonga below, the coaches are saying “vitesse, vitesse” which is french for “speed, speed”. He’s only doing 2 reps (with considerable load), but his aim is to throw the bar towards the ceiling (a cue we often use). This not only increases strength, but has been shown to increase power - and can set the stage for more power oriented training down the line (using lighter implements, med ball etc).

But if we use loads that are too light or perform too many reps with a given load, or don’t adhere to certain key guidelines, we won’t get the benefit of lifting (whether that’s during the bench or any other exercise that aims to increase strength and power).


Here are a few of those guidelines that you should adhere to if you aim to increase bench press (and upper-body) strength/power:

  • Resist the urge to ‘lift to failure’

  • Use rep schemes that don’t surpass 5 or 6 (but again, leaving 2-3 reps left in the tank)

  • Use higher set schemes than normal (up to 4-5 with novices and as much as 8 with players experienced in the weight room)

  • The intent to move the load must be very very high (even if the actual movement speed is low; remember that)

  • Leave sufficient time in between sets to feel ‘somewhat’ fresh for the next set

  • The upper-body will fatigue quicker than the lower-body, so you likely won’t be able to maintain the same loads and/or rep schemes on all sets. You can either progressively decrease the reps on each set (I like to do a 5, 4, 3, 2, 1 rep progression or similar) OR you can manipulate the loads (going up/down, or just down or doing a pyramid structure).

  • Combined with speed-specific serve sessions, benching may enhance both serve speed and quality (more on that below)

  • Even with younger teens, build up strength over time with similar principles

Other Factors to Consider + Final Word

There are other reasons we can bench press (or perform some other type of pressing action) - upper-body integrity is one of them - which can be helpful in high-speed movements, when changing direction and even when receiving high forehands (around shoulder level…which as you can probably recall, Agassi was quite good at).

Also, lifting in general can facilitate up-regulation of certain hormones (particularly testosterone) and better regulate others (cortisol). This is one reason I sometimes prescribe in-competition benching (but only single arm and only the non-dominant arm). Instead of using a lower-body lift that may contribute to local fatigue or using the dominant arm (which could be fatigued and in need of recovery from serving/hitting forehands), we can still get the hormonal benefit without any of the drawbacks (soreness, fatigue, pain etc).

In general, I prescribe a lot more dumbbell bench pressing work - usually easier to perform technically, you can get better range of motion (ROM) and thus more scapular activity, compared to the traditional barbell variation. But I still believe there is a time and a place (usually during a general prep block) where barbell benching can be implemented successfully - you simply cannot replicate the force expressions that are elicited with this lift, with any dumbbell alternative (one key factor that barbell benching could in theory, increase serve speed).

But you do have to earn your right to bench. That means that pulling activities (both horizontal and vertical) should be programmed concurrently. Some coaches and researchers advocate a 2:1 (or even 3:1) ratio of pulling to pressing. That may be a decent general rule of thumb, but not one that I necessarily follow. We do more pulling in our programs than pressing anyway (especially given that we do some type of Olympic lifting on a regular basis, which requires a lot of pulling...and with decent loads).

Lastly, as it stands today, there’s no definitive answer that yes you should or no you shouldn’t bench for tennis. We do it. Other players don’t. As you saw from above, you can increase throwing speed without increasing upper-body strength too. I believe the same is true in tennis. You can increase serve and forehand velocities without lifting...but lifting might help accelerate the process (and perhaps add a few other benefits in the meantime). But proper programming and execution cannot be stressed more, it’s a must! If you don’t have a coach to teach it well, it may not be the right fit for you or your players. My recommendation - do it well or don’t do it at all!


No fluff or spam - just a weekly post and the latest promos.



Member Login
Welcome, (First Name)!

Forgot? Show
Log In
Enter Member Area
My Profile Not a member? Sign up. Log Out