The purpose of this study was to compare the effects of recovery type (active vs. passive) on the time spent at VO(2)max and above 90 % of VO(2)max during short intermittent runs (15 s) at 120 % of maximal aerobic speed. Twelve male subjects performed a graded test and intermittent runs (15 s) alternated with 15 s of active recovery (50 % of maximal aerobic speed) or with 15 s of passive recovery to exhaustion. The time to exhaustion during the intermittent runs with active recovery (445 +/- 79 s) was significantly shorter (p < 0.001) compared with passive recovery (745 +/- 171 s). No significant difference was found between time spent at VO(2)max and above 90 % of VO(2)max for intermittent runs alternated with active recovery (180 +/- 121 s and 282 +/- 117 s, respectively ) and intermittent runs alternated with passive recovery (191 +/- 135 s and 317 +/- 132 s, respectively ). The times spent at VO(2)max (p < 0.05) and above 90 % of VO(2)max (p < 0.001), expressed in percentage of time to exhaustion, were significantly longer for intermittent runs alternated with active recovery (41 +/- 27 % and 64 +/- 24 %, respectively) than with passive recovery (25 +/- 16 % and 43 +/- 16%, respectively). In conclusion, the present study has shown that the time to exhaustion during repeated high-intensity exercise was significantly greater for passive compared with active recovery, despite no difference in absolute time spent at a high percentage of VO(2)max. However, the time spent at a high percentage of VO(2)max, expressed as a percentage of time to exhaustion, was significantly higher for active compared with passive recovery.