This randomized cross-over study aimed at comparing the recovery effect of 4 days of low-intensity, discipline-specific training of 1 vs 3 h daily. Eleven athletes completed two periods of 13 days intensive cycling training (IT), followed by a recovery period consisting of 4 days of low-intensity cycling for either 1 or 3 h each day. Before IT, after IT and after the recovery period, subjects were tested in the laboratory: venous blood sampling, "profile of mood states" (POMS), graded cycling test and a 30-min time trial (TT). Maximal heart rates and lactate concentrations decreased significantly after IT. Peak power output, maximal heart rates and maximal lactate concentrations changed significantly different during the recovery periods. Whereas these parameters were similar to pre-training values after 1-h daily active recovery, 3-h recovery training (REC) led to further decreases. Power output during TT was neither affected by IT nor by both recovery periods. TT-induced increases in cortisol, adrenocorticotropic hormone and prolactin were reduced only after 3-h REC. Total POMS and subscores fatigue and vigor changed significantly different during the recovery periods, a return to pre-training levels after 1 h active recovery and a further deterioration after 3 h REC. It is concluded that low-intensity training of a 1-h duration each day is more appropriate for recovery after an IT period than 3 h.