Four-day cycling rats were kept in a room with the lights on from 0500-1900 h. Injection of an ovulation-blocking dose of phenobarbital at 1230 h on proestrus did not alter the rise in plasma LH concentration in response to rapid injection of 12.4, 124, or 1240 ng of LHRH at 1300 h. In additional blocked rats, blood was rapidly withdrawen through one of two indwelling atrial cannulas while LHRH was infused at a constant rate through the other. Administration of a given amount of LHRH by slow infusion was much more effective in elevating plasma LH than was rapid injection of the same amount of releasing hormone. The pattern of plasma LH concentration after infusion of approximately 50 ng of LHRH per hour from 1500-1810 h was remarkably similar to that of the spontaneous LH surge, i.e., a gradual rise in plasma LH concentration followed by a steep linear increase to high levels which remained elevated for a period of approximately 2 h before declining rapidly soon after the end of infusion. In rats given a second 3 h infusion of LHRH at the same rate, from 2200 to 0100 h, a different response pattern was seen: the initial increase in plasma LH was greater, but the linear rise, which ensued after a lag period of similar duration (about 45 min), was less marked. Prolongation of the first infusion beyond 1800 h did not prolong the plateau: plasma LH levels declined before the infusion was terminated. This decline was less rapid than that seen at the end of a spontaneous LH surge, which in turn was less rapid than that seen after termination of a 3 h and 10 min infusion starting at the same time (1500 h), suggesting that both LHRH release and pituitary responsiveness are diminished (but not abolished) at this time. These findings clearly indicate that, in phenobarbital-blocked proestrous rats, the major part of the proestrous LH surge (including rising and plateau phases) can be simulated by a constant-rate infusion of about 150 ng of LHRH over the 3 h and 10 min period beginning at 1500 h. Since the data also suggest that LHRH release is reduced and/or occurs sporadically during the terminal phase of the surge (when plasma LH levels are declining), these experiments imply that the LH surge is caused by a properly timed, nearly constant-rate release of LHRH for about 3 h (beginning about 1500 h of proestrus) followed by a period of diminished LHRH release.