Previous data indicate that the infant rat shows a marked increase in adrenocortical responsiveness to stress immediately following prolonged maternal separation. In Experiment 1 we studied the immediate effects of repeated maternal deprivation. Our results indicate that the increase in basal as well as stress-induced corticosterone levels is a direct function of the length of the deprivation period immediately preceding testing, and is not cumulative. In Experiment 2 we examined the long-term consequences of maternal deprivation on adrenal responsivity. Four days following a single 24-h period of maternal deprivation, pups remained hyperresponsive to stress, although their basal levels of corticosterone had returned to control values. Shorter periods of deprivation (which did result in increased responsivity immediately following deprivation) did not have persistent effects. Our data suggest: 1) short periods of deprivation do not have a cumulative effect, and 2) there is a critical length of deprivation beyond which persistent changes in adrenocortical responsivity ensue.