The purpose of the present study was to examine whether the magnitude of the changes in the concentration of muscle metabolites influences the recovery of power output following short-term maximal intensity cycle exercise performed at different average pedalling rates. In part A of the study eight female subjects performed four trials on a cycle ergometer. Two trials involved maximal sprints of 30- and 6-s duration separated by a very short (2–3 s) recovery period. Average pedal rate during the first 30-s sprint was manipulated by employing resistances of either 7.5 or 10.1% of body weight; the second sprint always being performed against 7.5% BW. In two further trials subjects performed only a single 30-s sprint against the two resistances with pre- and post-exercise muscle biopsies and blood samples being taken. Peak power in the second sprint was significantly higher (442 ± 31W vs. 402 ± 33W; P < 0.05) following prior exercise against the greater resistance during which average pedal rate was lower (≈ 26%; P < 0.01) compared with the lesser resistance. However, despite this the muscle metabolite responses to the first sprint were similar (ΔPCr (7.5 vs. 10.1% applied resistance) –55 vs. –59 mmol kg dry muscle−1: ΔLactate + 104 vs. +107 mmol kg dry muscle−1: both P > 0.05). In part B of the study six female subjects performed 19 trials in which the recovery interval between a maximal 30-s sprint (where average pedalling rate was manipulated in a manner similar to part A) and a 6-s sprint ranged from 0 to 300 s. The rate of restoration of power output was influenced by the average pedal rate in sprint 1 only for recovery durations of up to 3 s. These findings suggest that the recovery of power is not exclusively determined by muscle metabolites, in particular PCr, when the recovery duration is very short (≤ 3 s). As it has been previously shown that the pattern of muscle activation influences ionic balance it is speculated that ionic factors may be very important in the early and rapid recovery of power.