How to safely train close to failure

Training to failure is effective, but is it more effective than non-failure training? Is training to failure really necessary? It turns out that in most cases, the benefits do not outweigh the risks of injury and overtraining. In this article we show you how to get the benefits of training to failure, without actually performing those last risky reps that lead to failure. Here’s how to safely train close to failure.

By Loek Vossen

Train to failure

Menu:

In some cases, training to failure could be beneficial to non-failure training. So before we talk about how to train close to failure, we should first answer the question: when should I train to failure, and when should I not. To do this, let’s first look at the meaning of training to failure.

What is training to failure

I’m sure you’ve heard of the well-known meaning of training to failure: complete as many reps as possible until you can’t complete another one. In other words: finish your set with no reps left in the tank, or no reps in reserve (RIR).

In this article we stick to this definition of repetition failure, but it’s good to know that there are actually more ways to train to failure.


Subscribe to the newsletter to get a notification when we publish the next article, that covers a different kind of muscle failure.


Rep failure can occur at the very first repetition, e.g. when you’re trying to lift a weight that is above your 1RM. But it could also be the 20th repetition in a low weight – high volume training.

You will intuitively understand that when a muscle fails to do its task, the body has an incentive to improve the performance of the muscle, to prevent failure next time. But is it necessary to train to failure?

Should you train to failure – according to the science

The discussion of whether you should train to failure or not has existed for a long time already. This makes sense, because training to failure has its pros and cons. Let’s summarize the most obvious ones.

Benefits of training to failure

Does training to failure increase strength?

As mentioned, it’s intuitively easy to understand that when a muscle fails to do its task, the body has an incentive to improve the performance of that muscle. It will therefore come as no surprise that training to failure is effective when you want to increase strength.

However, according to this meta-analysis (scientific review), you can expect similar increases in muscular strength when comparing training to failure with non-failure training.

Does training to failure increase hypertrophy?

How about muscle hypertrophy (increasing muscle size)?

Again we see that training to failure is effective when you want to increase muscle size. However, another scientific meta-analysis shows that you can expect similar results in muscle hypertrophy when comparing training to failure with non-failure training.

There seems to be one exception: resistance-trained athletes do significantly increase muscle size more when doing training to failure, compared to non-failure training. Since many of you are experienced athletes, this is an important finding. However, only 2 scientific studies were included in this analysis, which means more research is needed.

Drawbacks of training to failure

Can training to failure cause injuries?

Of course training to failure also has its downsides. For instance: because training to failure has a high training load, it can increase the chance of getting injured, according to this scientific review.

Does training to failure cause overtraining?

According to the same review, training to failure can also increase the chance of getting overtrained, when it’s performed repeatedly over long periods.

This makes sense, because when the training load is high, recovery between workouts becomes extra important. If you train to failure too often without enough recovery, you risk becoming overtrained.

On the other hand, when you do recover long between workouts, you maybe don’t train as often as you would, when not training to failure.

In the long run you might actually have a lower training load when training to rep failure.

Should you train reps until failure?

Looking at the pros and cons, it becomes clear that you can train to failure occasionally, especially when you’re an experienced athlete who wants to grow muscle. However, it also becomes clear that it doesn’t make sense to train to failure every training session. For some (power) athletes, training to failure is probably rarely (if ever) part of the workout.


As mentioned earlier in this article, there are more ways to train to failure, besides rep failure. There is one training to failure that I recommend doing every training. Leave your email address to get a notification when the article about that type of muscle failure is published.


Once we agree that training all the way until failure is not necessary, the next question is: how to safely stay away from failure?

How to train close to failure

If you want to have a high training load without the drawbacks of going to actual muscle failure, it makes sense to train close to failure. But how do you know whether you’re close to failure?

Some athletes and coaches will use metrics like:

  • Rate of Perceived Exertion (RPE)
  • Reps In Reserve (RIR)
  • Reps left in the tank
  • Reps until failure

If they were accurate, they would work perfectly fine. The problem is that athletes are not able to subjectively predict these metrics. In other words: they don’t know how many reps they can perform until failure.

According to this scientific meta-analysis, athletes underpredict the number of repetitions to task failure, regardless of how much experience they have in the gym. In practice that means that if a coach tells his/her athlete to leave 1 or 2 reps in reserve, they will leave (much) more reps in reserve.

The result is that all your training sessions are less challenging than prescribed.

The easy solution is to replace the subjective scores with something you can measure with a velocity tracker: (bar) velocity.

1RM velocity

For the sake of simplicity, I continue this article with the assumption that you’re trying to move a weight as fast as possible.

In a repetition that leads to failure, the speed of the weight becomes 0 before you finish the exercise. The weight is not moving anymore, you didn’t make it.

To succeed in a lift of maximum effort, you need a minimum velocity. The velocity at which you perform your last repetition before failure is called a minimum velocity. If you move below this velocity (while trying to go as fast as possible), you will not be able to finish the repetition.

According to this scientific paper, the velocity of the last repetition before failure is a given. It does not depend on the weight, and therefore the amount of reps. This is shown in the image:

Rep before failure velocity does not depend on weight
Whether you do more reps with a 60% of 1RM weight, or less reps with a 75% of 1RM weight, the last repetition before failure (marked with a blue circle) has a similar velocity (marked between 2 orange bars).

When you perform a 1RM, this rep is the last successful repetition before failure. The same rule applies here: the velocity of this last successful rep is a given. In other words, your 1RM velocity equals the velocity of your 6th rep with a 6RM weight.

This minimum velocity turns out to depend on the exercise, but remains roughly the same between athletes, regardless of their level. That means you already know your (bar) speed of your last successful rep.

You could derive these minimum velocities per exercise from the literature, e.g.:

  • Minimum velocity bench press: 0.15 m/s
  • Minimum velocity squat: 0.27 m/s

You could also easily test it yourself, by performing a set that leads to failure. As shown, this could be performed with a low weight (e.g. 60% of 1RM), all the way up to your 1RM weight. The velocity will be the same.

Using 1RM velocity to train close to failure

When you already know the velocity of your last successful repetition of that exercise, you can easily determine how far away you’re from failure.

After measuring the velocity of your first rep, you have an objective measure of how far away you are from failure:

Velocity and minimum velocity or 1RM velocity
When you measure your current velocity (blue bar) and you know your final rep velocity (which is a given: orange bar), then you know how far away you’re from failure.

In the example you lift a weight with a velocity of 1.5 m/s. You know the last rep velocity is a given, in this case 0.3 m/s. You’re now 1.2 m/s away from your last rep, which means you’re not approaching failure.

How to train to failure without a spotter

If you want to safely train close to failure without a spotter and without running the risk of injuries and overtraining, you can use the difference between your current velocity and the minimum velocity. This enables you to stop close to failure.

For example: if your bar speed during a bench press turns out to be 0.18 m/s and the minimum velocity is 0.15 m/s, you’re super close to failure. An extra repetition is probably not possible without a spotter. And as research has shown: this last repetition is probably not going to add any training effect anyway.

Reps In Reserve (RIR) velocity

According to this scientific paper, the last velocities close to failure don’t depend on the weight either. For instance: when your bench press velocity is in the range of 0.26 – 0.28 m/s you only have 2 reps in reserve. This turned out to be true regardless of the weight (e.g. 65% 1RM, 75% 1RM, 85% 1RM) and athlete experience (e.g.: novice, well-trained, highly-trained).

You could again derive these numbers from literature, or test them yourself, doing a set until failure.

By knowing the velocity of your 1RIR, 2RIR, 3RIR etc., you can prescribe how close to failure you want to train, using an objective measure.

More about knowing your reps in reserve by looking at the velocity in the next article. Subscribe to the newsletter to get a notification!

Loek Vossen

Loek Vossen

Human Movement Scientist | Content Marketing and Education