Why is recovery talked about as a half-life?

Because the decay of training fatigue is closer to exponential than to linear. In the first few hours after a session, a large fraction of the acute stress clears. Over the next day, more clears. By two half-lives, about 75 percent of the session's fatigue signal is gone. By three, about 87 percent. Using a half-life frames recovery in a way that handles both short and long recovery windows consistently.

What's a typical recovery half-life for a chest session?

Roughly 24 to 36 hours for a moderate session at RPE 7 to 8, in a well-conditioned lifter. Harder sessions (RPE 9+, novel exercises, high eccentric load) push that out to 36 to 60 hours. Smaller sessions at lower RPE can be under 24 hours.

Why do legs take so much longer to recover?

Three reasons stack. Leg movements (squats, deadlifts) produce more muscle damage per session because they involve larger muscle mass and more eccentric loading. They tax systemic recovery resources (glycogen, CNS fatigue, mechanical stress on the spine and hips) more than upper-body work. And the leg muscles themselves — especially quads — are large enough that local recovery mechanisms take longer even without the systemic piece.

Does my recovery half-life change over time?

Yes. It gets longer with age, with sleep debt, with underfueling, and with accumulated training load. It gets shorter when those factors improve. A well-calibrated system tracks the half-life over time rather than treating it as static. Short-term lengthening can also be a useful readiness signal — if your chest half-life is 20 percent longer than your personal baseline for the last two weeks, something in the recovery budget has shifted.

How do I measure my own recovery half-life?

You don't need to measure it directly. What you can observe is the pattern: does your next-day performance on the same muscle group match your expected ability, or does it lag? Does DOMS resolve by 48 hours or linger to 72? Does HRV rebound the morning after a leg session or stay suppressed? Consistent patterns across multiple sessions reveal the half-life without a lab.

Should I train a muscle before it's fully recovered?

Sometimes. The 'junk volume' concern assumes no training benefit from sub-recovery sessions, but research on frequency shows that training at 70 to 80 percent recovery can still produce adaptation and may be preferable to waiting for 100 percent, especially in smaller muscles. The deep risk is training at under 50 percent recovery, which tends to accumulate fatigue faster than it produces adaptation.

Per-Muscle Recovery Half-Lives: Why Legs ≠ Shoulders ≠ Chest

Your chest workout Monday felt great. By Wednesday, you barely felt like you’d trained. Then you hit legs Tuesday, and on Friday — 72 hours later — the stairs still felt wrong.

Why is one group bouncing back and the other dragging?

The answer is that your body doesn’t have one recovery timeline. It has many, and they differ by a factor of two to four across muscle groups. This post is about how per-muscle recovery half-lives work, why they differ, and what the pattern implies for training frequency and session design. It sits alongside the adaptive training pillar.

Recovery Isn’t Linear

A common mental model: “I’ll recover in 48 hours.” That’s a linear model — fatigue decreases at a constant rate until it hits zero.

Biologically, recovery is closer to exponential. The rate at which fatigue decreases is proportional to how much fatigue is still present. Right after a session, recovery is fast because the gradient is steep. A day later, recovery is slower because less fatigue remains. The mathematical shape is an exponential decay, and the useful summary statistic is the half-life: the time it takes for the fatigue signal to drop to half its initial value.

The practical difference between linear and exponential matters. With a half-life model:

At one half-life, you’re at 50 percent recovered.
At two half-lives, you’re at 75 percent (not 100).
At three half-lives, you’re at 87 percent.
At four, 94 percent.
At five, 97 percent — what most lifters would call “fully recovered.”

So a 24-hour half-life means you’re 75 percent recovered at 48 hours, 87 percent at 72, and approaching full recovery at 4 to 5 days. A 60-hour half-life (more typical for legs) means 75 percent at 5 days and 87 percent at about 7 days.

This framing also explains two confusions. First, lifters often think that if something takes “3 days to recover” it must be equally recovered throughout. It isn’t — most recovery happens early. Second, lifters sometimes think “fully recovered” is a discrete state, but physiologically it’s asymptotic: you approach baseline but only reach it after enough half-lives. At some point the residual is small enough to ignore, but “fully recovered” is a convention, not a physiological reality.

The Three Recovery Classes

Consumer training apps tend to use a single recovery constant per muscle group — 48 hours, maybe 72 for large muscles. The research evidence is more differentiated. Three broad classes cover most movements in a reasonable lifting program.

Class 1: Fast-recovering muscles (half-life roughly 20 to 30 hours). Calves, forearms, side and rear delts, and smaller upper-back muscles. These muscles are small, have shorter ranges of motion in their primary lifts, and produce proportionally less muscle damage per session. They can be trained three to five times a week in experienced lifters. The limiting factor is usually tendon and connective tissue, not muscle recovery.

Practical implication: the “calves twice a week” pattern that most lifters default to undertrains calves by a wide margin. High-frequency calf work (three to five sessions per week) is usually better. The same goes for rear delts.

Class 2: Medium-recovering muscles (half-life roughly 30 to 50 hours). Chest, biceps, triceps, mid-back. These are the muscles that support the traditional “every muscle twice a week” template. A Monday session is roughly 75 percent recovered by Wednesday and 87 to 90 percent by Friday, which makes the Monday/Thursday or Monday/Friday split physiologically sensible.

Harder sessions push the half-life up. An RPE 9 bench session with novel exercises might recover slower than a familiar RPE 7 session, because both the mechanical damage and the metabolic demand are higher.

Class 3: Slow-recovering muscles (half-life roughly 48 to 80 hours). Quads, hamstrings, glutes, lats, lower back. Large muscle mass, heavy eccentric loading in their primary lifts (squats, deadlifts, rows), and systemic recovery demand all contribute. A Monday heavy squat session is 50 percent recovered by Wednesday or Thursday and 75 percent by the weekend. This is why classical 3-day-per-week full-body programs work — they naturally leave 48 to 72 hours between leg sessions.

A lifter trying to squat and deadlift heavy three to four times a week is fighting the half-life, and the failure mode is usually accumulated fatigue showing up three or four weeks into the block.

Why the Classes Differ

Three mechanisms explain the variance.

Muscle damage per session. Larger muscles producing higher absolute force generate more micro-tears per session. The repair process is longer because more repair needs to happen. Eccentric-heavy movements (squat descent, deadlift lowering, hamstring curls) do more damage than concentric-dominant work.

Systemic recovery demand. A squat session pulls on far more than quadriceps. It depletes glycogen, stresses the spinal erectors, activates the central nervous system, elevates cortisol. Recovery requires replenishing all of that. A chest session hits a single muscle group and produces less systemic load. Same session volume in terms of reps, different systemic cost.

Connective tissue timeline. Tendons, ligaments, and fascia have their own recovery needs. They respond slower than muscle to training stimulus and detrain slower in absence of it. Muscles attached to long connective-tissue chains (quads via the patellar tendon, hamstrings via the sciatic chain, lats via the thoracolumbar fascia) inherit some of the connective tissue recovery timeline.

The combined effect produces the three-class pattern. Exceptions exist — some lifters have unusually fast leg recovery or unusually slow shoulder recovery based on body mechanics, fiber type distribution, and training history — but the population pattern is stable enough to use as a starting framework.

How the Half-Life Shifts With Context

The half-life isn’t fixed for an individual either. Several factors modulate it:

Sleep. Sleep debt lengthens the half-life across all muscle groups, legs most of all. A night of 5 hours of sleep can extend a chest session’s half-life from 30 to 45 hours and a squat session’s from 60 to 80.

Age. Biological age lengthens half-lives slowly but consistently. A 40-year-old’s half-lives are typically 20 to 30 percent longer than a 25-year-old’s for the same session. A 60-year-old’s are another 30 percent longer on top.

Life stress. Chronic stress — high cortisol, poor sleep, relationship or work stressors — lengthens half-lives. The effect is meaningful enough that experienced coaches can sometimes see it in the lift logs before the athlete mentions the life event.

Nutrition. Undereating, and specifically underfueling carbohydrates, lengthens recovery. Glycogen depletion combined with insufficient refueling can extend half-lives by 25 to 40 percent in the affected muscles. See energy availability for the systemic version of this.

Training age. Longer-trained lifters have shorter half-lives for the same session load than newer lifters. This is one of the benefits of consistency: your body gets better at recovering from familiar stimuli.

Novelty. A novel exercise or a novel rep range produces more damage and longer recovery than a familiar one. Adding a new variation on leg day spikes the next-day and next-next-day soreness in a way the regular movement doesn’t. After 3 to 4 exposures, the half-life shortens back to normal.

A system tracking these factors can update the half-life estimate continuously rather than treating it as static. A lifter whose chest half-life has been 32 hours for six months but is suddenly 48 hours in the current week is showing a signal worth interpreting — sleep debt, a nutrition change, or an accumulated-fatigue marker.

How Half-Lives Shape Training Frequency

A half-life model cleans up a debate that otherwise feels arbitrary: how often should you train a muscle?

The traditional answers are scattered. Hypertrophy research often suggests “2 to 3 times per week per muscle for optimal growth.” Powerlifting templates often use 3 to 4 times per week per muscle for compound lifts. Bodybuilding bro-splits train each muscle once per week. Concurrent training sometimes dictates 5 to 6 sessions per muscle per week.

All of these can be sensible, and the half-life framing unifies them.

The core principle: training a muscle when it’s at roughly 70 to 90 percent recovered is usually productive. Training at under 50 percent recovery is usually counterproductive. Waiting until 100 percent before the next session leaves adaptive stimulus on the table.

Given a muscle’s half-life, you can read off the appropriate frequency:

A 24-hour half-life means 75 percent recovered at 48 hours. Training every 48 hours (three to four times per week) is sustainable.
A 36-hour half-life means 75 percent recovered at 72 hours. Training every 72 hours (roughly twice per week) is sustainable.
A 60-hour half-life means 75 percent recovered at 120 hours. Training every 5 days (about one to two times per week) is sustainable.

A typical bodybuilder’s split — chest Monday, back Tuesday, legs Wednesday, shoulders Thursday, arms Friday — implicitly assumes weekly frequency, which over-recovers the fast classes. A typical Upper/Lower split trains each muscle twice per week, which matches the medium-class half-life but over-recovers the fast muscles and under-recovers none of them.

The half-life framing suggests the optimal frequency is different for different muscles in the same program. Calves three to five times a week. Chest and back twice. Quads and hamstrings one to two times. Running a single frequency across all muscles means either the fast muscles are undertrained or the slow ones are overtrained.

Within-Session Half-Lives: Inter-Set Recovery

A subtler application of the half-life framing is within a session. Rest between sets is effectively a partial recovery process, with a much shorter half-life.

For a set of 5 at near-failure, the metabolic recovery half-life between sets is roughly 90 to 180 seconds. Three minutes of rest puts you at about 75 percent metabolic recovery. Five minutes puts you near 87 percent. For heavy strength work, this explains why 3 to 5 minute rests produce better output than 60 to 90 second rests.

For bodybuilding-style training in the 8 to 15 rep range, the metabolic load per set is higher and the recovery is slower, but the downstream stimulus is less sensitive to complete recovery. Rest of 60 to 120 seconds produces lots of stimulus from a growth standpoint and is the typical default.

The key insight: within-session rest scales the same way as between-session rest. It’s the same exponential decay, just at a different timescale. A lifter who rushes rest between sets is essentially training at an even lower recovery percentage than the nominal session load suggests.

Reading Half-Lives In Your Own Data

You don’t need lab equipment to observe per-muscle half-lives in your own training. What you can track:

Next-session performance. A session that’s well-recovered produces similar or better top-set performance than the previous week. A session attempted too early produces noticeably slower bar speed, missed reps, or dropped top-set loads.

DOMS timeline. Delayed onset muscle soreness typically peaks 24 to 48 hours post-session. If your chest DOMS is gone by 36 hours, your chest half-life is probably 20 to 30 hours. If your quad DOMS is still present at 72 hours, quad half-life is over 50.

HRV trend. Overnight HRV usually dips after hard sessions and returns to baseline. The dip-and-return pattern tracks the systemic half-life, which for big compound sessions is usually a combination of the muscles involved and the CNS load.

Subjective readiness. A daily readiness check-in, tracked over months, reveals patterns. Wednesday morning readiness after Monday chest sessions vs Thursday morning readiness after Tuesday leg sessions. If the second pattern is consistently worse, the leg half-life is longer than the chest half-life.

A well-designed adaptive system aggregates these signals into a single per-muscle half-life estimate, updates it weekly, and uses it both in the MAV prescription and in day-to-day session planning.

Implications for Session Order and Split Design

Once half-lives are explicit, a few program design decisions that feel like preference become easier to reason about.

Heaviest sessions first. When two muscles will be trained back-to-back (say, legs Tuesday and an accessory back session Wednesday), the longer half-life should go first in the week so it gets more recovery time before the next same-muscle session.

Don’t stack two high-half-life muscles close together. Hard legs Monday and hard deadlifts Tuesday overlap in systemic recovery demand. The second session accumulates on top of the first. Two days of separation is often better.

Use high-frequency patterns for fast-recovery muscles. If calves or rear delts are priorities, distribute them across the week rather than bundling them into a “shoulders and calves” day. The bundle under-trains the muscle; the distribution maximizes stimulus given the short half-life.

Let deloads map to slow-recovery muscles. A deload week benefits slow-recovery muscles most. Chest and back recover quickly enough that a deload is mostly a mental reset for them. Legs and lower back genuinely need the deload volume reduction.

These aren’t inflexible rules — programs are personal and a good coach can justify exceptions for any of them. They are defaults that become reasonable when the half-life framing is explicit.

What the System Can’t See

A honest list of what half-life modeling doesn’t capture:

Technical recovery. Motor learning and skill retention recover faster than muscle. A lifter can re-execute a technique 48 hours after a session even if the muscle isn’t at full capacity.

Psychological recovery. Heavy sessions take a mental toll that’s hard to quantify. A lifter can be physiologically recovered but psychologically drained after a string of RPE 9 sessions. This shows up eventually in subjective markers but the time course isn’t a half-life.

Injury-specific recovery. If a session produced an actual injury — a strain, a tendon flare, a joint irritation — the recovery timeline is governed by tissue healing, not the normal half-life. Treating injury recovery with a session-level half-life model produces bad output.

Acute illness. A cold, flu, or gastro event extends recovery across all half-lives. Modeling this as a half-life shift is possible but the illness itself is a bigger signal that needs separate handling.

In Summary

Per-muscle recovery half-lives are:

Faster in small muscles (calves, shoulders, forearms) — 20 to 30 hours
Medium in torso muscles (chest, biceps, triceps) — 30 to 50 hours
Slower in large muscles (quads, hamstrings, lats, back) — 48 to 80 hours
Responsive to sleep, age, stress, nutrition, and training age
Different from the generic “48 to 72 hours” that most apps use

A training system that treats all muscles as having the same recovery window is leaving information on the table. The adaptive training intelligence stack uses per-muscle half-lives as one of several signals, combined with MAV ceilings, RPE calibration, and ACWR/monotony/strain context.

The prescription layer prescribes. The recovery layer times. Together, they produce a week that your body’s actual recovery dynamics can support — not the textbook’s.

More on Omnio’s implementation: /features/adaptive-training.