This study investigated the effects of a moderate (MI) and a low intensity (LI) active recovery (both compared to a passive recovery) on repeated-sprint performance and muscle metabolism. Nine, male, subjects performed three repeated-sprint cycle tests (6 x 4 s sprints, every 25 s) in a semi-randomized, counter-balanced order. Recovery after each sprint for the MI and LI trials, respectively, was 60 W (approximately 35% VO(2max)) and 20 W (approximately 20% VO(2max). Biopsies were taken from the vastus lateralis pre- and immediately post-test during the MI and LI trials to determine adenosine triphosphate (ATP), phosphocreatine (PCr) and lactate (MLa(-)) content. Compared to passive, significant reductions in peak power of 3.4-6.0% were recorded in the MI trial (4 of 6 sprints; P < 0.05) and reductions of 3.5-3.7% in the LI trial (2 of 6 sprints; P < 0.05), with no differences between the two active trials. No significant differences were evident in ATP, PCr and MLa(-) between the two active recovery trials. In summary, peak power indices during the repeated-sprint test were inferior in the MI and LI active recovery trials, compared to passive. The minimal differences in performance and muscle metabolites between the MI and LI trials suggest that any low-to-moderate level of muscle activation will attenuate the resynthesis of PCr and the recovery of power output during repeated short-sprint exercise.