Training max guide

Start with a current estimate

Use a recent set from the same lift, same equipment, and same technique standard.

The more stale or sloppy the input is, the more conservative the training max should be.

• Use a recent hard set.
• Avoid old meet PRs as current training numbers.
• Keep technique standards consistent.

Rank the input before ranking yourself

The best training max starts from the lift you are actually about to train. A paused bench, touch-and-go bench, low-bar squat, high-bar squat, strapped deadlift, and mixed-grip deadlift can all produce different numbers.

When two inputs disagree, use the one that best matches the next block and then add the buffer. Do not average a clean current set with a stale all-time PR just because the bigger number feels nicer.

• Best input: recent clean set from the same lift, equipment, stance, grip, range, and rep standard.
• Okay input: recent e1RM from a similar setup, with a larger buffer.
• Weak input: old PR, different variation, post-peak meet result, or a grindy set with changed technique.

Choose the buffer on purpose

If your estimated 1RM is 180 lb, a training max might be 160 to 165 lb for a conservative block, or a little higher when the estimate is recent, clean, and the program expects that setup.

The buffer is not weakness. It is what keeps percentage work from turning into accidental max testing.

• Use a smaller buffer only when the estimate is recent and the lift is stable.
• Use a larger buffer after time off, technique changes, poor sleep, dieting, or missed sessions.
• Use a larger buffer when the program repeats hard percentage work for several weeks.

Keep the three numbers separate

Your estimated max, training max, and one-day work load each have a different job. The estimate describes what you might be able to lift now, the training max lowers that number for planning, and the work load is what you actually put on the bar that day.

Do not feed the lowered training max back into a 1RM calculator as if it were your true strength. Do not treat one lighter backup load as a new training max either. Keep the labels separate so the next decision is based on the right signal.

• Estimate or test strength only from a lift that matches the next block.
• Lower that estimate into a training max for percentage work.
• Choose daily work loads from the training max and the session target.
• Record one-day backups as session adjustments, not permanent max changes.

Let the program calculate from the training max

Once you choose the training max, percentages come from that number, not from your all-time best lift.

This makes the early weeks feel manageable and gives you room to accumulate good reps.

• Calculate percentages from the training max, not the all-time PR.
• Round loads to the smallest practical plate jump.
• Keep early weeks clean enough that later weeks still have room.

Compare like with like before judging it

A training max can look wrong when the real change was the training context. Before you raise or reset it, check whether the lift standard, range of motion, equipment, stance, grip, rest periods, assistance work, bodyweight, or recovery target changed.

If the comparison is messy, keep the next edit small. Hold the training max, standardize the setup, or repeat the exposure before treating the number as too light or too heavy.

• Same lift variation, same range, and same rep standard.
• Same rest target and similar warm-up quality.
• Similar assistance work, conditioning, food, sleep, and bodyweight trend.
• If context changed, collect one cleaner comparison before rewriting the max.

Write a backup load before you need it

Before the session starts, write the planned work weight and the lighter load you will use if warm-ups, first reps, or technique quality clearly disagree with the spreadsheet.

The backup is not a punishment or a new training max. It is a one-day adjustment that keeps the session inside the intended effort while you collect better evidence.

• Plan the work weight from the training max.
• Pick a small rounded-down backup load before warm-ups begin.
• Use the backup when warm-ups are unusually slow, reps lose their normal shape, or the first work set overshoots the target effort.
• Keep the training max unchanged until the same pattern repeats.

Change it only when the trend is clear

One great day does not require a big jump. One bad day does not always require a reset.

Look for repeated evidence: clean reps, stable recovery, and work sets landing near the intended effort.

• Do not raise the max from one unusually easy set.
• Do not reset the max from one bad session.
• Look for several clean or several grinding exposures.
• Use RPE, bar speed, and recovery as context.

Adjust the work before rewriting the max

If one session is off, change the day first: trim the load, stop a set early, or keep the next set cleaner.

Only change the training max when the same signal repeats across several exposures. The training max should follow the trend, not the mood of one workout.

• One unusually easy day: keep the max and collect another clean exposure.
• One rough day: adjust that session and check sleep, warm-ups, pain, and fatigue.
• Several clean weeks: consider a small increase.
• Several grinder weeks: hold, reduce, deload, or reset the training max.

Review it at the cycle boundary

At the end of a block, sort the signal before you change the number. Ask whether the main lifts were too easy, repeatable, or repeatedly above the intended effort.

If the work was clean and recovery stayed normal, add the smallest planned jump the program expects. If the first sets were fine but later sets collapsed, fix rest, assistance, or set count before blaming the training max. If warm-ups and first work sets were already grinders for several similar weeks, reset the training max and rebuild.

• Clean block: small planned increase.
• Main lift fine, week overloaded: trim assistance or total set count first.
• First work sets repeatedly too hard: lower the training max for the next cycle.
• Pain, technique collapse, or major life stress: do not use the training max as the only explanation.

Estimated max to training max

A lifter estimates a 140 kg squat from a recent hard set. They choose a 90% training max, so their training max is 126 kg.

A 75% work set would then be based on 126 kg, not 140 kg.

If the first week calls for 5 reps at 75%, they load about 95 kg and judge whether the set matches the intended effort.

Two inputs, one block

A lifter has a 150 kg all-time deadlift from last year, but their current best clean set estimates closer to 140 kg after a technique change.

For the next block, they set the training max from 140 kg because it matches the pull they are actually practicing.

If the new setup stays clean for several weeks, the training max can move up. If they start from the old 150 kg PR, every percentage set is already pretending the technique change did not happen.

When to lower it

If three weeks of work sets all land above the intended RPE and technique gets worse, the training max is probably too high for the current block.

Lowering it can preserve productive training better than forcing missed reps.

For example, a 126 kg squat training max might drop to 120 kg for the next cycle if 75-85% work keeps becoming grinders instead of repeatable work.

When not to raise it

A lifter hits one unusually fast top set after great sleep and extra caffeine. That is useful information, but it is not enough to rewrite the whole block.

They keep the training max steady, repeat the next week, and only increase if the clean performance keeps showing up.

A conservative increase might be 2.5 kg for upper-body lifts or 5 kg for lower-body lifts after several clean weeks, not a jump from a single good day.

End-of-cycle review

A lifter finishes four squat weeks. The first work set was usually clean, but the last backoff set kept drifting above the target effort after heavy leg presses and short rests.

That is not automatically a bad training max. For the next cycle, they keep the training max steady, trim one overlapping accessory slot, and use longer rests before reducing the main squat number.

If the first work set also becomes a grinder for several similar weeks, then the training max gets lowered instead of asking the rest of the week to rescue a bad input.

Planned load, backup load

A lifter has a 126 kg squat training max. The day calls for 80%, so the planned work weight is about 100 kg after rounding.

Before warming up, they write 95 kg as the backup. If 100 kg moves normally, they train as planned. If warm-ups are slow and the first rep at 100 kg already feels like the target effort for the whole set, they use 95 kg for the remaining work and leave the training max alone.

If that same backup choice is needed for several similar squat sessions, then the cycle review has real evidence for lowering the next training max.

Three numbers, three jobs

A lifter estimates a current bench 1RM around 110 kg. They choose a 100 kg training max for the block, then calculate an 80% day from 100 kg, so the planned work load is 80 kg.

If warm-ups are unusually slow, they might use 77.5 kg that day. The log should not say their new max is 77.5 kg, and the next 1RM estimate should not be built from the 100 kg training max. The estimate, training max, and daily adjustment stay in their own boxes.

Same number, different comparison

A lifter switches from touch-and-go bench to paused bench, adds two extra close-grip sets, and shortens rest because the session is running late. The same 80% now feels much harder.

That does not prove the training max is suddenly wrong. The cleaner first move is to compare the paused bench under stable rest and assistance conditions, then decide whether the training max, the new variation, or the added work needs the adjustment.

What if the first week feels too light?

That is often the point. A training max should leave room for clean reps, repeatable practice, and later weeks that get heavier.

Before raising it, check whether the planned effort stays too easy for several similar sessions. If it does, make a small planned increase next cycle instead of turning one easy week into a max test.

Should I use 90% because my program says training max?

Only if the program actually expects that setup and your estimate is current. Some programs define a training max as a specific percentage; others use the phrase more loosely.

Read the program instructions first, then sanity-check the number against the first work sets. If the early sets are already grinders, the input or buffer is probably too aggressive for this block.

Can I test a true max while using a training max?

You can, but testing and training have different jobs. A true max test should be planned, recovered for, technically consistent, and not forced into the middle of a block just because the training max feels boring.

If you test, use the result as new information. Do not automatically rebuild every work set from it unless the next block is designed around that tested lift, range of motion, and recovery context.

Progression depends on the lifter

ACSM resistance-training guidance emphasizes progressive training with variables such as load, volume, frequency, exercise choice, rest, effort, and training status.

That is why the same estimated max may need different training maxes for a beginner, a tired intermediate, and a peaking powerlifter.

Percentages are estimates

Shimano and colleagues found that reps at a given percent of 1RM vary by exercise, and a later meta-regression also reported wide variation in repetitions completed at the same percentage of 1RM.

That is why percentage tables should be treated as planning estimates. A training max protects the program from normal estimate error, lift-specific differences, and day-to-day performance swings.

Autoregulation keeps the number honest

RPE and RIR research gives lifters a way to compare planned load against actual effort.

If a training max repeatedly creates the wrong effort, update the training max instead of worshiping the spreadsheet.

Session changes are weaker evidence than trends

A one-day backup load is a practical response to the known uncertainty in percentage estimates and daily effort ratings.

It can protect the intent of the session, but it should not be treated as proof that the whole training max is wrong until the same signal repeats.

Layered numbers need clean labels

Research on repetitions at percentages of 1RM and RPE-based loading supports using estimates and effort feedback with margins of error.

It does not mean every lowered planning number becomes a new strength estimate. Keeping the estimate, training max, and daily adjustment separate prevents one cautious layer from being counted twice.

Context changes weaken the signal

Percentage and effort research supports checking load against the actual lift, lifter, and session response. It does not make a changed grip, new pause, shorter rest, extra assistance work, or different recovery state equivalent to the old comparison.

When several variables changed at once, the training max is only one suspect. Standardizing the comparison first makes the next adjustment less random.

Cycle reviews need the whole week

A training max is only one variable inside the week. Load, volume, rest, exercise choice, assistance work, and recovery can all change how the same percentage feels.

That is why a cycle review should separate an overestimated max from too much later-set fatigue or a recovery problem before the lifter resets the number.