Muscle phosphocreatine (PCr) recovery time constant (an index of muscle aerobic capacity) and contractile ATP cost were estimated from a gated (31)P NMR protocol which does not require intense, repetitive exercise. Subjects performed 2-s duration, maximum voluntary isometric ankle dorsiflexion contractions at 30-s intervals for 8 min (total 15 contractions), while single-shot (31)P spectra (51.7 MHz, TR 3 s) were acquired from the anterior compartment muscle. Spectra from the sixth through 15th contractions were retrospectively sorted, yielding 10 spectra (each 10 averages) gated to times before and after contraction. There was no significant decrease in muscle pH, allowing the calculation of contractile ATP cost directly from the percentage change in PCr during contraction cycles [8.86 +/- 0.82% (SE, n = 11) of PCr at rest], corresponding to an ATP cost of 1.69 +/- 0.16 mM/s (range 0.99-2.49 mM/s), assuming an 8.2 mM ATP concentration. The time constant for PCr recovery (tau 41.8 +/- 4.2 s, range 22.0-60.8 s) was calculated from tau = -Deltat/ln[D/(D + Q)], where Q is the percentage change in PCr due to contraction, D is the additional steady-state percentage drop in PCr from rest and Deltat is the interval between contractions. In the same subjects, the monoexponential PCr recovery time constant after more intense, repetitive isometric ankle dorsiflexion exercise (30 s at 0.5 Hz, 50% duty cycle) was similar to (36.2 +/- 3.5 s, range 16.5-58.8 s) and well correlated with (r = 0.82) the gated result. In contrast to the gated protocol, muscle pH decreased from 7.01 +/- 0.01 to 6.78 +/- 0.04 during recovery after the repetitive protocol. Hence the gated protocol allows the estimation of muscle ATP cost and PCr recovery without intense exercise or muscle acidification.