How to deload: when to take one, what to cut and what to keep

The most common deload mistake is treating it as a week off. Knowing how to deload starts with the opposite idea: you keep training on the same days with the same lifts, and you cut the sets and the effort. The app's planned deload week halves your working sets, targets 4.5 reps in reserve, and drops the next weight by about 10 per cent. For a lifter squatting 132.5 kg for six, that week is two sets at 120 kg, and nothing in the program adds load until it is over.

A deload is a short, planned stretch of easier training meant to shed fatigue so the next block starts fresh. The expert definition is close to that: a period of reduced training stress designed to ease physical and mental fatigue and prepare you for the training that follows [1]. RIR (reps in reserve) is how many more reps you could have done when you racked the bar, and RPE is the same scale read the other way, so RPE equals 10 minus RIR.

Cut sets and effort first

Volume is the one thing every expert agrees should come down. In a Delphi study of strength and physique coaches, 34 coaches answered the first round and 21 finished the third, and the panel agreed universally that training volume should drop during a deload [1]. Lowering sets and stopping further from failure both drew full agreement as ways to do it. Lowering reps inside a set was close behind.

Lifters do the same thing when nobody is watching. A survey of 246 competitive strength and physique athletes found every one of them deloaded [2]. They cut repetitions per set and sets per week, raised their reps in reserve, and lifted lighter loads.

The app's deload week follows that pattern with two fixed numbers. The volume multiplier is 0.5, half the sets you would normally do. The effort target is 4.5 reps in reserve, which is RPE 5.5. Those numbers come from the app's own block code, and they are the same whether the deload ends a hypertrophy, strength or accumulation block.

Load is the third lever, and the gentlest. The app drops the next weight by about 10 per cent on a planned deload week, rounded to a weight you can load on the bar. It does not halve the load, because a deload is meant to keep the movement familiar. An empty-bar week teaches your legs nothing about 130 kg.

What to keep the same

Keep the exercises. The coaches' panel agreed that exercise selection could stay unchanged during a deload, and that main lifts could be trained with lower volume or effort rather than dropped [1]. The surveyed athletes mostly kept the same exercises too [2]. New movements in a recovery week risk new soreness, which is the opposite of the point.

Keep the days. Athletes in the survey reported that training frequency stayed the same through the deload [2]. The coaches agreed that frequency need not change unless life is busy, and that sessions get shorter on their own because there is less to do [1]. A 75-minute leg day becomes a 40-minute one, and you still turn up on Tuesday.

The warm-up stays too. A deload set at 120 kg still deserves a ramp, and the app's warm-up builder writes the same one it would for any working weight of 120 kg: the bar for 8, then 60 kg for 5, 85 kg for 3 and 102.5 kg for 2. Warm-ups never count toward volume, so the ramp costs nothing on the weekly set count.

Log every set as usual. A deload week is a week of data like any other, and the session after it is easier to judge when the numbers on either side are written down.

A deload week, worked through

Take a 90 kg lifter on a block that has been going well. Their last normal squat session was four working sets of 132.5 kg for 6 at RPE 8. The model behind the one-rep max calculator puts that set at an estimated 168.6 kg for one.

The plan then reaches its deload week. Here is what the app writes for that lift, and for the resensitize option covered below.

The deload load is 90 per cent of 132.5 kg, rounded to a loadable 120 kg. Half of four working sets is two. On the same model, the load this lifter could move for six at RPE 6 is 124.6 kg, so six reps at 120 kg sits below even that. The whole week should feel like holding back.

Log the RPE you actually felt, with one caution. The model reads RPE from 6 upward. Anything below 6 is treated as RPE 10, the same assumption it makes for a set logged with no RPE at all. Log those six reps at 120 kg as RPE 5.5 and the estimate for that set reads 143.4 kg, well short of the 168.6 kg the lifter really has. That is a deload set read as a max. It does not mean anything has been lost, and the post on why a lift goes down explains how to read one low number against the noise.

The app also refuses to add load anywhere in that week. A deload week overrides every other signal, including a stall flag and a cut, and never emits an increase.

When to plan one, and when to react

Plan one at the end of each block, and bring it forward when the bar tells you to. That is what the research and the surveys point to, and it is how the app schedules them.

The consensus paper notes that deloads are generally taken every 4 to 6 weeks and last about 7 days, though some run from a single session to two weeks [1]. The surveyed athletes reported a typical deload of 6.4 days every 5.6 weeks [2]. In the app, any working block of four weeks or more ends on a deload week automatically. The training blocks guide shows where that week falls in each of the seven block types.

Coaches on the panel agreed that a deload can be pre-planned, autoregulated, or both, and that how often depends on how a lifter responds to training [1]. The athletes named the triggers for an unplanned one: performance stalling, more muscle soreness than usual, and joint aches [2].

The app has its own reactive version, at the level of a single lift. By default, three sessions without progress at the same load flags a stall and holds the weight. If the stall carries on for another session, or the estimated 1RM is trending down, the app suggests backing that lift off by about 10 per cent and rebuilding. That suggestion is never written into the plan for you. It is a nudge to deload one lift rather than a whole week.

A rough guide for deciding which you need:

• One lift stuck for four sessions while the rest move: back that lift off, keep the week as planned.
• Several lifts flat, sleep poor, joints sore: take the deload week now rather than waiting for the planned one.
• Everything moving and the block has a week to run: finish the block, then deload as planned.

The resensitize option

Sometimes the load is fine and only the volume is the problem. After a long, high-volume push, a lifter may still be strong at their working weights but no longer growing from the extra sets. For that, the app offers a Resensitize block.

A resensitize week holds the load, trims volume to 0.6x, and asks for 3 reps in reserve. The app's note for it is plain: keep the weight easy and stop well short of failure. For the 90 kg lifter, that means 132.5 kg stays on the bar for two or three working sets. On the model, 132.5 kg at RPE 7 is a set of five, so each set stops one rep earlier than the normal week did. Like a deload, a resensitize week never adds load.

The idea behind it is that a stretch of lower volume makes later volume work again. The evidence for that is limited. In one trial, 14 young men benched 3 sets of 10 at 75 per cent of their one-rep max, three days a week [3]. One group trained straight through for 24 weeks. The other trained in 6-week cycles with 3 weeks off between them. Both groups ended with similar gains in muscle size and strength, and the periodic group grew faster than the continuous group during its second retraining cycle.

That study used weeks off, not a reduced week, and the bench press alone. It says breaks did not cost muscle over six months. It does not prove a 0.6x week will make your next block grow more.

What the evidence can and cannot say

The honest summary is that deloads are universal in practice and thin in the trials. The only controlled trial of a one-week deload in trained lifters we could find is recent [4]. Thirty-nine resistance-trained men and women ran a nine-week, high-volume program. One group stopped training for a week at the midpoint, and the other trained straight through.

Muscle growth, power and local endurance came out the same in both groups [4]. The group that kept training gained more lower-body strength, and reported slightly better readiness to train. So a week off did not cost muscle, and it did cost a little strength.

A week off is not what most programs mean by a deload, which is why the app keeps you training through it. Two sets at 120 kg keeps the squat pattern practised. Nobody has yet run the trial comparing that against training straight through, so treat the deload as fatigue management with a low cost, not as a growth tool. The coaches' panel itself noted that regular pre-planned deloads may not be necessary for everyone, and that some coaches skip them [1].

Plan the week, run it as written, and judge it by the session after. If the first week of the next block moves well, the deload did its job. If you would rather the app wrote those weeks for you, that is what Top Sets does, and coming from another logger the switching guide covers moving your history across.

Questions

Should I take a full week off instead of a deload?

You can, and you probably will not lose muscle. In the one trial of trained lifters, a week off mid-program matched continuous training for muscle growth, but the continuous group gained more strength [4]. A reduced week keeps your technique practised and your schedule intact, which is why the app plans a lighter week rather than a blank one. Take the week off if life demands it.

How long should a deload last?

About a week. The consensus paper notes deloads generally run for around 7 days, with some as short as one session and others as long as two weeks [1]. Competitive athletes in the survey averaged 6.4 days [2]. The app's planned deload is one week at the end of a working block of four weeks or more.

Do I lower the weight or the sets on a deload?

Mostly the sets. Every expert on the Delphi panel agreed volume should come down, and cutting sets and stopping further from failure both drew full agreement [1]. Load can drop a little. The app halves your sets, targets 4.5 reps in reserve, and drops the next weight by about 10 per cent, so you still handle a familiar load.