REF-034 · Black et al., 2015 · Applied Physiology, Nutrition & Metabolism, Vol. · Paper

Self-Pacing Increases Critical Power

Blimited evidenceMeasured on recreational to trained athletesWritten before October 2026, being re-checked

Participants

n=10 males (age 21.5 ± 1.9 years; mass 75.2 ± 11.5 kg). Each completed 3-4 constant work-rate (CWR) tests and 3-4 self-paced time-trial (TT) prediction trials on a cycle ergometer for the determination of CP and W'.

Protocol

Within-subject comparison on a cycle ergometer: ten males each completed 3-4 constant work-rate (CWR) prediction trials and 3-4 self-paced time-trial (TT) prediction trials, and CP and W' were derived from each protocol (from the abstract).
Measures: CP and W' by protocol; mean response time of pulmonary O2 uptake; total O2 consumed in the first 60 s, correlated with CP across both protocols (n = 20 data points).

Overview

Black et al. tested whether critical power (CP) derived from a series of self-paced time-trials is higher than CP derived from the conventional series of constant work-rate tests. It was: TT-derived CP 265 ± 44 W vs CWR-derived CP 250 ± 47 W (P < 0.05), a difference of about 15 W (~6%, arithmetic on the two means).

W' did not differ between protocols (TT 18.1 ± 5.7 kJ vs CWR 20.6 ± 7.4 kJ, P > 0.05). The mean response time of pulmonary O2 uptake was shorter in the TTs (34 ± 16 vs 39 ± 19 s, P < 0.05), and CP correlated with the total O2 consumed in the first 60 s across both protocols (r = 0.88, 20 data points).

The authors conclude that a self-selected pacing strategy enhances CP and severe-intensity exercise performance, so TT completion times may be overestimated by CP and W' parameters derived from CWR protocols.

Key Findings

Self-paced (TT) CP: 265 ± 44 W; constant work-rate CP: 250 ± 47 W; P < 0.05
Difference ~15 W (~6% higher with self-pacing) — arithmetic on the two means
W' not different between protocols (18.1 ± 5.7 vs 20.6 ± 7.4 kJ, P > 0.05)
Faster VO2 kinetics in the TTs: mean response time 34 ± 16 vs 39 ± 19 s (P < 0.05)

Practical Application

In 10 young males, CP from self-paced time-trials was 265 ± 44 W versus 250 ± 47 W from constant work-rate tests (~6% higher, P < 0.05), while W' was unchanged — the testing modality moves the CP estimate, so record how a CP/FTP value was obtained and compare like with like.
Not from this paper; the base's own practice: if racing is self-paced (most road racing), derive CP/FTP from self-paced efforts; if the target is a steady time-trial, a constant-load protocol is closer to the demand. Document the method so a shift between test types is not misread as a fitness change.
Inference, not tested: TT completion times predicted from CWR-derived CP and W' may be pessimistic (the authors' own suggestion), so pacing plans built on constant-load CP can leave power on the table in a self-paced race.

Limitations

n=10 young males; training status not reported in the abstract, so the "trained cyclists" label used earlier was unsupported.
Laboratory ergometer; neither CP value was validated against race performance.
How this entry was checked and changed
Why this level

regraded 2026-09-27 from B+: within-subject protocol comparison, n=10

Content correction 2026-09-27
Lead line now states the study's finding (265 vs 250 W, n=10) instead of a testing procedure; the unsourced "constant-load 20-min or 30-min MLSS" advice was removed and the remaining testing advice is labelled as the base's own practice.
Tier: "trained cyclists" -> Tier 1-2; the abstract describes ten young males and gives no training status. Level B+ -> B (within-subject protocol comparison, n=10).
Added W' and VO2-kinetics results from the abstract.
2026-09-28: added Protocol from the abstract (Europe PMC PMID 26088158).