Ss (N=72) were randomly assigned to four groups and performed two bouts of either 1 min. of sustained and/or 3 min. of rhythmic isometric grip-flexion exercise. The two activity bouts within each testing session were separated by a recovery interval of 5, 10, 20, 40, 80, 160, or 320 sec. Every S was tested following all recovery intervals. The criterion for assessing recovery was the summation of force exerted at each 5-sec. interval during the sustained task or at each 15-sec. interval during the rhythmic exercise. During a rhythmic second activity bout significantly greater force-time integrals were developed subsequent to an initial bout of rhythmic activity. During a sustained second bout, however, there was no difference between the groups exposed to the two types of initial exercise. In both cases significantly greater force-time integrals were exerted subsequent to longer inter-bout recovery intervals. The patterns of recovery appropriate to the four possible permutations of exercise were described capably by two-component, double-positive exponential equations.