Training to failure has long been debated in bodybuilding circles, with some swearing by all-out intensity while others advocate leaving reps in reserve.
But what does science actually say about how hard you need to push for maximum muscle growth?
A recent deep dive into training intensity examined real-world examples from elite bodybuilders alongside scientific research to answer this critical question.
The findings might surprise anyone who thinks every set needs to be an all-out war.
Jump to:
What Failure Actually Means
According to research from Fischer and Steele, muscular failure occurs when despite maximal effort, a person cannot meet exercise demands, causing involuntary set termination. Your muscles physically can’t produce enough force to move the weight.
Coaches typically break this into two categories: absolute failure and technical failure. Absolute failure means you can’t move weight even with compromised form. Technical failure happens when you can’t complete another rep with proper technique.
Most coaches prefer technical failure because it feels wrong to define success as pushing past safe movement patterns. However, this creates problems too—some form deviation naturally occurs as you approach failure, especially with free weights.
The sweet spot? Allow minor form deviation to reach muscular failure, but stop before creating an entirely different exercise. On machines where form stays consistent, tolerate less deviation.
Rep Speed Reveals Hidden Truth
Researchers have started investigating failure through rep speed slowdown. As you near failure, concentrics move more slowly during grinding final reps.
A 2019 study found bench press reps slowed 30% halfway through sets, then moved 80% slower on final reps compared to initial ones.
Individual variation matters significantly—different exercises produce different slowdown patterns. Still, if reps don’t slow at all, you’re likely not approaching failure. Noticeable slowdown suggests you’re getting close.
Two Schools of Thought
The Every-Set-To-Failure Approach
Some lifters take every set to complete failure, eliminating guesswork about effort levels. This isn’t just meathead training—peer-reviewed literature supports training to highest intensity for recruiting maximum motor units and muscle fibers.
However, recommendations always exist within low-volume, single-set-to-failure paradigms. This approach works for time-constrained individuals seeking efficient stimulus with minimal time investment.
The problem? You must keep weekly volume extremely low when taking multiple sets to failure, and substantial research shows higher volume typically produces more muscle growth.
The Volume-Based Alternative
A meta-analysis led by James Krieger examined 15 studies—11 favored higher volumes for muscle growth. When broken down by sets per muscle weekly, doing at least 10 sets clearly beats 5-9 sets or fewer than 5 sets.
A systematic review and meta-analysis from Gergich and colleagues analyzed 15 studies with roughly 400 participants. Results showed similar muscle size increases regardless of whether training reached muscular failure.
Other research confirms those final two pre-failure reps create disproportionately more fatigue. This recovery demand varies—lateral raises to failure differ vastly from deadlifts—but going to failure consistently demands more recovery per set.
RPE and RIR: The Better System
Rating of Perceived Exertion (RPE) ranks set difficulty from 1-10. Reps in Reserve (RIR) quantifies remaining reps. RPE 10 equals failure with zero RIR. RPE 9 means one rep remained. RPE 8 indicates two reps left in reserve.
Expert opinions vary on optimal proximity to failure:
- Dr. Mike Zourdos suggests 5-7 reps from failure (RPE 3-5) works in certain contexts
- Dr. Eric Helms recommends 0-5 reps in reserve (RPE 5-10) for hypertrophy
- Dr. Brad Schoenfeld advocates 1-3 reps in reserve (RPE 7-9), taking final sets to failure on select exercises
Context determines appropriate effort. Sufficiently heavy weight allows leaving 4-5 reps in reserve because you activate most motor units from rep one, especially as advanced trainees. Lighter weight exceeding 12-15 reps requires staying within 1-2 reps of failure.
The Subjective Problem
RPE and RIR rely heavily on subjective assessment. Poor estimation risks falling below effective thresholds.
Solution: perform anchoring sets. Choose exercises permitting safe failure, call out when you think two reps remain, then push to true RPE 10. This calibrates your perception against reality.
Recording sets provides another calibration method. Watch playback and observe rep speed changes. If reps don’t slow noticeably, you’re probably not pushing as hard as perceived, even if effort feels high mentally.
Real-World Examples Analyzed
Chris Bumstead’s set demonstrated true RPE 10. Rep five took 1.27 seconds, rep six 1.23 seconds, rep seven jumped to 1.9 seconds. Rep eight held at 1.9 seconds, rep nine slowed to 2.5 seconds, and his final rep ground out at 3.2 seconds with spotter assistance.
I personally would save this level of exertion for the last set of an exercise or you could even stop two reps shy of where Chris did to mitigate fatigue a bit.
John Meadows demonstrated excellent one-rep-in-reserve execution. Explosive tempo and consistent speed throughout, then hitting the wall abruptly on final reps. Based on last rep speed, he could manage one more grinder but nothing beyond.
Kai Greene’s eight-rep set showed RPE 7-8 territory—two to three reps remaining. Initial reps moved smoothly and consistently. Final rep speed indicated another 2-3 reps possible, making this perfectly stimulative given the heavy load.
When Effort Falls Short
Sets terminated at RPE 5 or lower typically prove insufficiently stimulative for maximizing hypertrophy. No rep speed slowdown, no grinding reps—clear indicators of excessive reps remaining.
Alberto Nuñez’s pulldown set contrasted this perfectly. His final three reps showed technical mastery with serious grinding—the last concentric required 4.17 seconds. Perfect effort for final back work.
Individual Variation Matters
Steffi Cohen’s deadlift set demonstrated why judging RIR from single reps fails. Her first rep suggested immediate failure, yet she completed ten total reps with the final one grinding nearly five seconds.
Big individual differences exist—you can’t always isolate single reps to determine remaining capacity.
Sunny Andrews’ elevated dumbbell squats showed another complication. Using strong mind-muscle connection, constant tension, and smooth tempo makes rep speed judgment harder. Lifters emphasizing controlled cadence over explosive tempo often make weight look easy until hitting the wall abruptly.
The Practical Takeaway
Maximum hypertrophy doesn’t require every set reaching absolute failure. Research clearly supports this, and elite examples confirm it.
Most training should land between RPE 7-9 (one to three reps in reserve), adjusting based on load, exercise selection, and individual response. Reserve true failure for final sets of training blocks, safer exercises, or occasional anchoring sessions.
Record your training, study rep speed changes, and calibrate perceived effort against actual performance. This develops the subjective skill of effort assessment while maximizing volume and managing fatigue—the true keys to long-term muscle growth.
