Taper
calculator
A pre-race volume-reduction schedule from your peak training week. Exponential fast-decay (the meta-analytic winner) against the linear step-down most guides print — and the one lever they get wrong.
| Taper week | Exponential | Linear |
|---|
—
How the math works
A taper is a progressive reduction of training volume in the days before a race, to shed accumulated fatigue while keeping the fitness you already built. The primary literature is specific: maintain training intensity, hold session frequency (cut it by no more than ~20%), and take the reduction out of volume (Mujika & Padilla 2003). The meta-analysis (Bosquet et al. 2007, 27 of 182 studies) found the most efficient version is a roughly 2-week taper along an exponential fast-decay curve, with volume cut 41–60% and intensity and frequency left alone.
Set the race-week target as a fraction of your peak week, then walk each taper week down to it. This page shows both time-courses for the same endpoint:
N = taper length in weeks, i = 1..N (week N = race week)
exponential (fast-decay): volume_i = peak × f^(i / N) ← Bosquet 2007's winner
linear (even step-down): volume_i = peak × (1 − (1 − f) × i / N)
Both hit peak × f in the race week; the exponential cuts more early.
Frequency floor (Mujika 2003): keep at least ceil(0.80 × sessions).
Volume is unit-agnostic — miles, kilometres, or weekly hours all scale the same way, because the schedule is a pure fraction of your peak.
Worked example
A runner peaks at 40 mi/week (64.4 km) and takes the Bosquet optimum: a 2-week taper, exponential fast-decay, 50% cut at race week (f = 0.5), pace and session count held the same.
- Week 1 (2 weeks out): 40 × 0.5^(1/2) = 28.3 mi (45.5 km) — 70.7% of peak. The linear step-down would run 30 mi (48.3 km), so the exponential front-loads the cut.
- Week 2 (race week): 40 × 0.5 = 20 mi (32.2 km) — 50% of peak, both curves.
- Frequency: five sessions stays five (or four at the most); the 20-mi race week is ~4 mi (6.4 km) per run, not fewer runs. An effective taper improves endurance performance ~3% (range 0.5–6.0%, Mujika & Padilla 2003).
When this calculator is wrong
The taper's benefit is well quantified — about 3% on endurance performance, usual range 0.5–6.0% (Mujika & Padilla 2003) — and it comes from a specific recipe: a 2-week, exponential, 41–60% cut in volume, with intensity and frequency left alone (Bosquet et al. 2007, effect 0.72 ± 0.36 for the volume band). The common failure is cutting the wrong knob.
- Cutting intensity. Ease off the pace and you detrain the top-end fitness you spent months building; the schedule here cuts distance, not effort. Keep the hard sessions hard.
- Cutting frequency too far. Thinning a 5-day week to 3 days is a 40% frequency cut — past Mujika's ~20% floor — even if the weekly mileage lands on target. Keep the session count (or drop one at most) and shorten each run.
- Wrong starting point. If you arrive at the taper overreached, sick, or injured, the 3% model does not apply: you need actual rest, and pulling intensity and sessions is then the right call. The floor is guidance for a healthy taper, not a reason to train through a red flag.
What to do with the result
Use the race-week number as the ceiling for your final week and the exponential column as the target for each week between now and the race. Hold your workout paces where they were at peak; the sessions just get shorter and, if anything, a touch more spaced. Keep your usual number of training days — take the reduction out of the length of each session, not the count. If your peak week is in hours rather than miles, read the percentages instead of the volumes: the schedule is the same shape either way. Recompute if your race date moves or your peak week changes.
Common questions
- How long should a taper be?
- The meta-analysis found a roughly 2-week taper the most efficient (Bosquet et al. 2007); the broader range in the literature is about 4 to 28 days (Mujika & Padilla 2003), shorter for shorter events and less-trained athletes, longer for marathons and high-volume athletes.
- How much should I cut my volume?
- By 41–60% at the race-week low point (Bosquet et al. 2007). This page defaults to 50%, the midpoint. Cutting less can leave residual fatigue; cutting more risks losing fitness without extra freshness.
- Should I run slower during the taper?
- No. Intensity should stay the same — reducing it causes detraining and removes most of the benefit (Mujika & Padilla 2003). The workouts get shorter, not easier.
- Should I run fewer days per week?
- Keep your session count, or cut it by no more than ~20% (Mujika & Padilla 2003). Dropping from five days to three is a 40% cut and starts to detrain, even if you hit the same mileage. Shorten each run instead.
- Why does the exponential schedule differ from the linear one?
- Both reach the same race-week volume, but the exponential fast-decay curve takes more off early and levels out into race week. Bosquet et al. (2007) found that shape more efficient than an even step-down for the same endpoint.
- How much faster will a taper make me?
- About 3% on endurance performance, with a usual range of 0.5–6.0% (Mujika & Padilla 2003) — a few minutes over a marathon. It is a real but bounded effect, not a leap in fitness.