The 5/3/1 Workout Program Explained

Quick answer: The 5/3/1 workout program is a strength template by Jim Wendler built on four barbell lifts: overhead press, deadlift, bench press, and squat. Every working weight comes from a training max set at 90 percent of your true one-rep max. Each cycle runs three weeks of rising percentages plus a deload, and the last set of every session is an AMRAP.

You wrote 5/3/1 at the top of a page in your notes app three weeks ago. Now you are staring at a number like 102.5 and trying to remember whether it came off a 135 training max or a 130 one.

That is the tax on percentage programs. The lifting is simple; the bookkeeping is where people quit.

What is the 5/3/1 workout program?

Four lifts, one per session, four sessions a week. Wendler's 5/3/1 is built around the overhead press, deadlift, bench press, and squat. Each day opens with one of those lifts for three working sets, and everything after that is accessory work. The name comes from the rep targets rotating across three weeks: sets of 5, then sets of 3, then a 5/3/1 week.

The program is deliberately slow. You add a small amount to each lift every three weeks, roughly 10kg (20lb) a year on the press and 20kg (40lb) on the squat. That sounds disappointing until you compare it to the alternative, which for most stalled lifters is zero.

What is a training max, and why 90 percent?

Every percentage in 5/3/1 comes off a training max, not your real max. You set the training max (TM) at 90 percent of your true one-rep max on that lift. If your squat 1RM is 150kg (330lb), your TM is 135kg (297lb), and every prescribed weight is a percentage of 135.

The 10 percent haircut is the safety margin. Wendler's site frames the training max as a way to control your programming rather than a scoreboard, and puts it plainly: a higher training max does not equal more strength. Lifters who inflate it grind reps and stall inside two cycles.

You do not need to test a true 1RM to start. Estimate a max from a hard set you have already done using our one rep max calculator, then take 90 percent of it. Our guide to estimated 1RM covers how much to trust that number. A too-light TM costs you one cycle; a too-heavy one costs you three.

What are the 5/3/1 percentages week by week?

Three working sets per lift, at rising percentages of the training max. Here is the standard scheme, with + marking the all-out set:

Warm-ups sit below all of this and do not count. The deload has no plus set: you do the reps written, you leave, and you start the next cycle rested. Our guide to the deload week explains why planned easy weeks are worth the patience. Wendler's later books make that fourth week optional, but the three-week wave above has not changed.

The plus set carries the program

That last set is not a formality. On 85% x 5+ you do 5 reps as a minimum, then keep going for as many clean reps as you can. You stop when bar speed drops or form degrades, not at absolute failure.

It also audits your training max. Five reps at 95 percent in week 3 means the TM is honest. One grinding rep means it is not.

That audit matters because a percentage is an average and you are not one. A meta-regression of 269 studies covering 7,289 individuals found that at 80 percent of 1RM, people averaged 13.1 reps on the leg press but only 8.8 on the bench press, with wide individual spread on top of that.

The plus set autoregulates in a way fixed percentages cannot. In an 8-week randomized trial in Frontiers in Physiology, 21 trained men lifted either on fixed percentages of a pre-test 1RM or on self-selected loads targeting an RPE. Strength gains came out statistically similar. Percentages get you close; effort on the day settles the rest.

How do you progress between cycles?

You add weight to the training max, not to the bar. After each three-week wave, raise the TM by 2.5kg (5lb) on bench and overhead press, and 5kg (10lb) on squat and deadlift, then recalculate every percentage from the new number. The bar gets heavier because the number it derives from got heavier, which is a different mechanism from the session-by-session jumps of a linear progression.

When to stall the increase: if you cannot beat the minimum reps on the plus set, leave that lift's TM alone. If you missed badly, cut it by 10 percent and rebuild. Wendler is blunt with beginners on this: if the reps are not clean, you do not increase the training max.

Who should run 5/3/1, and who should not?

5/3/1 for beginners is usually the wrong call. A raw novice can add 2.5kg to the squat every session for months. Trading that for 5kg every three weeks is a bad deal. If linear progression still works for you, keep taking the free strength.

5/3/1 suits the lifter who just stopped getting it. Six months to two years in, your last few attempts at adding weight ended in missed reps, and you are tired of resetting. The light weeks and the slow TM climb turn "add weight until you fail" into something you can run for a year. Lifters coming off a stalled 5x5 land here often.

What assistance work goes with 5/3/1?

The three main sets are fixed. Everything after them is a template choice. Boring But Big (BBB) is the best known: after your 5/3/1 sets you do 5 sets of 10 on the same lift (or its opposite) at 50 to 60 percent of the training max, then some pulling and core work.

Other templates swap that volume for bodyweight work or a few accessory lifts. The rule underneath all of them is that assistance should not compromise the main lift next session. For a second opinion on how yours balances out, get a free private link to the workout plan optimizer and paste your week in.

How do you track a percentage-based program?

5/3/1 is unusually punishing to run from memory. A rep-based program only asks what you lifted last time. This one asks for four training maxes, which cycle week you are in, and how many reps you hit on the last plus set. Lose any of those and the next session is a guess.

A simple starting plan

1. Find a recent hard set on each of the four lifts. One set of 5 you remember clearly is enough.
2. Set each training max at 90 percent of the estimated 1RM. Use the one rep max calculator, then round down to the nearest plate jump.
3. Write the four training maxes at the top of your log and date them. Every weight for three weeks comes from those numbers.
4. Log the plus set reps every session. Squat 128x8 is the line that decides your next cycle.
5. At the end of week 3, add 2.5kg upper and 5kg lower to any lift that beat its minimum reps, then start again.

Where does Gym Note Plus fit in?

Your notes app is fine for the lifting part. Typing three lines after a squat session beats tapping through any app's set logger, and you already do it. The gap is that plain notes cannot answer "what was my bench TM in April" without scrolling.

Gym Note Plus keeps the typing and adds the memory. You write the same lines you already write:

• Type Squat 101x5, 115x3, 128x8 and it logs three sets, including the 8-rep AMRAP, without forms.
• Your plus-set reps become a chart, so you can see whether 95 percent is getting easier across cycles.
• Old numbers surface when you search a lift, which is how you check a training max mid-session.
• Any template logs the same way, so a BBB block of Squat 67 5x10 sits right under the main work.

The percentages in 5/3/1 are just arithmetic. The program only works if the training max is honest and the plus set reps get written down.

Frequently Asked Questions

Is 5/3/1 good for beginners?

5/3/1 is usually not the best first program. Beginners can add weight every session on a linear progression, which builds strength far faster than the 2.5kg to 5kg per three-week cycle that 5/3/1 delivers. Switch once linear progression stops working and you are repeatedly missing reps.

What is a training max in 5/3/1?

A training max in 5/3/1 is a working number set at 90 percent of your true one-rep max, and every prescribed percentage is calculated from it. The 10 percent buffer keeps heavy sets makeable on bad days. Wendler advises going lighter if you are unsure, because an inflated training max stalls the program.

What are the 5/3/1 percentages?

The 5/3/1 percentages, all taken from the training max, are: week 1, 65 percent x5, 75 percent x5, 85 percent x5+; week 2, 70 percent x3, 80 percent x3, 90 percent x3+; week 3, 75 percent x5, 85 percent x3, 95 percent x1+. The optional deload uses 40, 50, and 60 percent for 5 reps each.

What does the plus sign mean in 5/3/1?

The plus sign in 5/3/1 marks an AMRAP set: do the listed reps as a minimum, then continue for as many technically clean reps as possible. Wendler advises stopping when bar speed drops or form breaks rather than going to absolute failure. Those rep counts tell you whether your training max is set correctly.

How many days a week is 5/3/1?

Standard 5/3/1 runs four days a week, one main lift per day: overhead press, deadlift, bench press, and squat. Wendler also publishes three-day and two-day versions where the four lifts rotate through fewer sessions, stretching each cycle over more calendar time. The four-day layout is the original default.

Can you build muscle on 5/3/1?

Yes, with a high-volume template. The three main sets per session are strength work and low in volume, so muscle growth on 5/3/1 comes mostly from assistance such as Boring But Big, which adds 5 sets of 10 at 50 to 60 percent of the training max after the main lift.

Final Takeaways

• Every weight comes off the training max, which sits at 90 percent of your true 1RM. Get that number wrong and the whole cycle is wrong.
• Learn the wave: 5s, 3s, then 5/3/1, at 65/75/85, 70/80/90, and 75/85/95 percent, with an optional deload at 40/50/60.
• Treat the plus set as the feedback loop. The same percentage allows very different rep counts between people and lifts, so what you hit on the last set is the signal worth reading.
• Progress the training max, not the bar: 2.5kg (5lb) upper body, 5kg (10lb) lower body, once per cycle. Hold any lift that missed its minimum reps.
• Track four training maxes and every AMRAP count. In a notebook, a notes app, or Gym Note Plus, a percentage program with no record of its own inputs is guesswork.