[en] Neuromuscular electrical stimulation (ES) and passive recovery (PR) were compared in ten healthy men after a provocation exercise inducing delayed onset of muscle soreness (DOMS). The exercise consisted of 3 sets of 30 maximal eccentric contractions performed by the knee flexor muscles of the dominant leg on an isokinetic dynamometer at 60 degrees/s angular velocity. There was an interval of 8 weeks between both bouts and the order of the recovery mode (ES or PR) was block-randomly assigned. ES recovery consisted of a 25-min continuous and non-tetanic (5 Hz) stimulation of the hamstring muscles. Concentric and eccentric hamstrings peak torques were evaluated before and immediately after the provocation exercise, after the recovery period, as well as 24 h (d1), 48 h (d2), 72 h (0) and 168 h (0) after the bout. Subjective perception of muscle soreness (VAS, 0-10 a.u.) was evaluated before exercise and at d1, d2, 0 and d7. To assess the CK activity, five blood samples were drawn before exercise and at d1, d2, d3 and d7. For both recovery modes, the greatest reductions in isokinefic muscle performances were measured on d2 (66.3 +/- 24.1 % of initial values (ES) vs. 57.4 +/- 26.5% (PR) for the concentric mode and 55.6 +/- 16% (ES) vs. 53.1 +/- 19.3% (PR) for the eccentric mode). d2 also corresponded to the highest painful sensations (5.4 +/- 2.14 a.u. (ES) vs. 6.15 +/- 2.55 a.u. (PR)). Peak activities of CK were reached on d3 (47507 +/- 19973 IU/l (ES) vs. 75887 41962 IU/l (PR)). Serum CK was lower with ES than PR at 0 (p <= 0.05) but all other parameters changed in a manner that was not statistically different between the two recovery protocols (p > 0.05). This strong trend could be explained by an electro-induced hyperperfusion that may efficiently wash out the muscle from the cellular debris resulting from the initial injury, and hence diminish the inflammatory response and the delayed amplification of tissue damages.