Cool-down and recovery between hard sessions
A short easy spin, food and sleep, a genuinely easy day next, and a gap set by what the session damaged; the gaps in hours and days are practice, and cold water after lifting costs gains
The question it answers
What to do in the hour after a hard session, what the next day should be, how long before the next key session after intervals, a long ride or run, a race or a heavy lift — and which recovery tools help without blunting the adaptation the session was for.
Default (what to prescribe first)
Cool-down: optional, 5-15 min easy, then eat and get warm. Against a passive cool-down, an active one was largely ineffective for same-day and next-day performance (some studies found next-day benefits), soreness, damage markers and injury; it clears blood lactate faster, not necessarily muscle lactate, may speed cardiovascular and respiratory recovery, and preliminary evidence says it does not blunt adaptation (REF-913, B+, narrative review, no dose).
Not from a study; the base's own practice: a comfortable spin or jog of 5-15 min that lets breathing settle. Cold water's benefit is thermal, with limited evidence that it clears metabolites (REF-737, B+, mechanistic review), so "flushing lactate" is not the reason to cool down. Refuelling is SYNTH-69.
The day after a key session is easy or off (not from a study; the base's own practice, SYNTH-28). Easy = comfortably below the first threshold, not pinned to it: in 20 trained amateur runners, easy running AT the threshold gave the same 6-week gains as 10-15% below it, with a higher psychophysiological load at some time points (REF-705, B+; running only).
Easy work is cheap: HRV was back to baseline 5-10 min after up to 120 min below VT1, versus ~30 min after threshold or 6 × 3-min work in 9 highly trained runners and ≥90 min after that interval session in 8 trained runners (REF-161, B; acute HRV, not adaptation).
Sleep is the recovery tool with no trade-off. Across 69 studies, ≤6 h of sleep cut performance ~7.6%, the harm coming from total deprivation and waking early, and afternoon/evening sessions suffered more in the pooled data (PM −8.31% vs AM −5.42%; endurance tasks were impaired at either time (AM −6.50%, PM −3.56%; not formally compared — the authors read it as irrespective of time of day)) (REF-204, A-) (corrected 2026-10-12).
The easy day — caps (consistent with [SYNTH-51])
Anchor to a measured LT1/VT1 when there is one; otherwise these house defaults, not findings:
Decision — gap to the next hard session
All gaps are house practice unless a refId is attached to that number.
| After | Next 24-48 h (practice unless cited) | Next key session (practice) | Source and how sure |
|---|---|---|---|
| VO2max or threshold intervals | Easy or off | ≥48 h; well-trained riders on a planned block may go back-to-back, then recover | REF-161 B (acute cost equal for threshold and VO2max work in the highly trained); a Giro podium rider never trained hard >2 days running and added easy days as load rose (REF-080, C+, n=1, Tier 5 — transfer); five 30/15 sessions in 6 days then 6 easy days raised power at VO2max 3.7% vs 0.7% (REF-048, B, non-randomised, well-trained cyclists) |
| Long ride below VT1 (≤ ~3 h) | Easy, or a normal day | 24-48 h | REF-161 tested ≤120 min only; one 6 × 10-min moderate-intensity interval session left more next-morning fatigue and soreness than the same work split in two (REF-159, B, n=14 runners — transfer to a long easy ride) |
| Long run / hilly or descent-heavy session | Easy; no sprints | 48-72 h | 3 days × 2.5 h: runners 133% more CK, 87% more soreness than cyclists (REF-235, B+); after 40 min downhill running leg stiffness stayed up to 96 h (REF-317, B+, 18 men) |
| 5 km-10 km, half marathon | Easy running from day 1-2 | ~3-5 days (5-10 km), ~7-10 days (half) | No entry — practice only |
| Marathon | Off, then easy 40 min from 48 h | ~2 weeks | Easy runs at 48/96/144 h did not slow CK/LDH recovery vs rest; squat jump better at 96 h (108.29% vs 100.58%) (REF-761, B+, n=64, mostly men). CK back by 144 h, LDH by 192 h, CRP still up at 192 h (REF-762, B, uncontrolled). The 2 weeks is SYNTH-64 practice |
| Mountain / 100 km+ ultra | Off, then easy | ~2 weeks, gated by soreness and RPE | Leg force −35-39%, back by 16 days, mostly by 9 (REF-662, B, 166 km mountain race, 11 of 22 followed up); SYNTH-65, SYNTH-61 practice |
| Cycling race, crit, gravel | Easy; no sprint or FTP test | 48-72 h | Muscle damage cut sprint and incremental peak power, not TT output; RPE and ventilation rose at the same VO2 (REF-701, A; after a damaging protocol, not a race — transfer); no cycling race-recovery entry |
| Stage race / camp | 2-3 easy days | when markers are back | LSCT values back to baseline 3 days after an 8-day camp (REF-755, C+, 6 pro women) |
| Heavy lower-body lift | Easy ride fine | ≥24 h; ≥48 h if heavy, eccentric or plyometric | SYNTH-31 (hours = practice on REF-329's direction, B+); same day ≥3 h apart (REF-330, A, strength-side outcome); hamstring stiffness peaked 48 h, back at 96 h while tightness felt worst at 24 h (REF-322, B) |
| Two-a-day | — | see below | — |
Masters. Widen gaps when monitoring says so; age alone is weak grounds — any age effect on recovery is smaller than assumed and largely training status (REF-598, B, narrative review, no hours). Keep the intensity (SYNTH-35).
Two-a-days. Strength and endurance ≥3 h apart (REF-330); the priority session first (SYNTH-31). With 4 h between two runs (the tested gap; "under ~6 h" is the entry's practice line), fructose plus maltodextrin gave a ~32% longer second run to exhaustion than glucose plus maltodextrin at the same total (REF-734, B+, n=8, running) — dose in SYNTH-69.
A 20-90-min nap between 13:00 and 16:00 with 30 min awake before training (REF-608, B, systematic review). Do not split a durability session: two halves 6.5 h apart reset the drift the long session was for (REF-159). Keep the cool-down short and eat first: an active cool-down can interfere with muscle glycogen resynthesis (REF-913).
No cold water in the gap before a hard effort (REF-737); partial immersion cut jump power immediately (g −0.94, REF-250).
Recovery modalities — help vs blunt
The "Use" column is the base's own practice built on the cited rows.
| Modality | What it does | Adaptation cost | Use |
|---|---|---|---|
| Easy movement / active recovery | Cut soreness (REF-385, A, 99 studies; corrected 2026-10-12); easy running after a marathon did not slow recovery (REF-761); as a cool-down, largely ineffective for recovery markers (REF-913, B+) | none shown (REF-913: preliminary) | Default |
| Massage | Most effective technique for soreness and perceived fatigue (REF-385, A); a 16-min massage did not change objective stiffness vs sham (REF-317, B+) | not tested in the base | For comfort; any time |
| Compression | Moderate benefit to soreness, strength, power, CK after damage (g 0.40-0.49); strength recovery largest at 2-8 h and >24 h (REF-387, A); in competitive and elite endurance athletes it did not improve sport-specific performance (SMD 0.04, moderate certainty, REF-1090, A-) | not tested in the base | After races, long or damaging days, into the next day |
| Cold water, 10-15 °C, 5-15 min | Lower CK and soreness at 24 h, no faster strength or jump recovery; legs-only = whole-body (REF-250, A); works mainly by cooling (REF-737); protocol from REF-386 (B+, narrative); whole-body immersion raised early-night slow-wave sleep vs partial-body in 12 well-trained runners, with no difference in fatigue or damage markers over 48 h (REF-019, B) | After strength sessions, 12 weeks of CWI after each session blunted strength and muscle gains (REF-011, B+, 21 active men); across eight controlled training studies regular CWI at ≤ 15 °C reduced strength gains (SMD -0.60) but did not change time-trial or maximal aerobic power (SMD -0.07) or time-trial duration (SMD 0.00) after endurance training, intervals wide (REF-914, A-) | When hot (REF-737) and between race days (practice); after endurance-only sessions no aerobic cost shown (REF-914); not after lifting in a build; not shortly before a hard effort (REF-737) |
| Foam rolling | After exercise eased muscle pain (+6.0%) and slightly protected sprint and strength; effects "minor" (REF-379, A) | not tested in the base | Optional |
| Static stretching, electrostimulation | No change in soreness (REF-385) | — | Not as recovery |
| Sauna / hot water after training | Performance in heat trivial and uncertain, prediction interval includes a loss (REF-334b, A-) | — | Not as recovery |
| NSAIDs | In an 80-km ultra, ibuprofen went with kidney injury in 52% vs 34% (REF-892, A-; interval includes zero) | no adaptation trial in the base | Not for soreness; medical use only |
| High-dose vitamin C + E | 1000 mg C + 235 mg E daily through 11 weeks of endurance training: VO2max +8% in both groups (REF-908, A-, 54 young adults) | Muscle COX4 +59% on placebo vs −13% on vitamins: markers blunted, performance not | Not in a build (practice; SYNTH-67); vitamins from food were not tested |
| Sleep, naps | REF-204; sleep loss impairs endurance and strength, extension helps (REF-203, C+, narrative) | none | Always |
How to progress / periodise
When NOT to use it / red flags
What the evidence does and does not show
Evidence gaps
How this entry was checked and changed
(modality effects on soreness rest on A meta-analyses, cold water and training adaptation on an A- meta-analysis, the cool-down on a B+ narrative review; race-recovery time courses on B/B+ running studies; the spacing between sessions — the core decision — is house practice over narrative reviews, a C+ single-rider case study and acute B crossovers)