Studies have shown that peripheral levels of corticosterone correlate with the magnitudes of two well-described physiological models of memory, long-term potentiation (LTP) and primed burst (PB) potentiation. In the present experiments, the authors investigated the effects of experimenter-controlled manipulations of the levels of corticosterone on the magnitude of hippocampal PB potentiation in urethane-anesthetized rats. Primed burst potentiation is a long-lasting (at least 30 minutes) increase in the amplitude of the CA1 population spike and EPSP slope in response to physiologically patterned stimulation of the hippocampal commissure. The levels of serum corticosterone were controlled by implanting corticosterone pellets in adrenalectomized rats (ADX/PELLET). In the first experiment, a significant negative linear correlation between elevated (stress) levels of serum corticosterone (greater than 20 micrograms/dL) and the magnitude of PB potentiation in ADX/PELLET subjects (r = 0.60, P < .05) was found. In the second experiment, the shape of the corticosterone-PB potentiation function was different at low and intermediate levels of corticosterone than it was at high levels of corticosterone: There was a positive correlation at low levels (0-10 micrograms/dL), a peak response at intermediate levels (11-20 micrograms/dL), and a negative correlation at high levels (21-93 micrograms/dL) of corticosterone. Thus, the overall relationship between corticosterone and PB potentiation is an inverted-U function. These findings provide strong support for the hypothesis that corticosterone exerts a concentration-dependent biphasic influence on the expression of hippocampal plasticity.