Reduced food intake results in altered reproductive function in female mammals and previous studies have shown this to be due, at least in part, to an increase in the negative feedback efficacy of estrogen on pituitary luteinizing hormone (LH) secretion. This study attempted to elucidate the site (CNS vs. pituitary) of increased negative feedback efficacy of 17β-estradiol (E<sub>2</sub>) in underfed female rats. Rats (Holtzman) were fed either ad libitum (CR) or 50% of normal food intake (UR). All other conditions were identical for CR and UR. When UR stopped exhibiting vaginal cycles, they were ovariectomized (OVX) along with the CR exhibiting normal vaginal cycles, and implanted subcutaneously with Silastic capsules containing 0, 1.2, 2.4 or 4.8 µg E<sub>2</sub>. Six or 12 days after OVX, rats were cannulated in the left carotid artery under brief ether anesthesia. Starting 3 h after surgery, blood was collected sequentially through a peristaltic pump at 5-min collection intervals for 2 h. Sequential plasma samples were assayed for LH by RIA. Pulse analysis was performed using PC PULSAR [ 15]. No significant differences in pulse frequency or amplitude were found between CR and UR either at day 6 or 12. However, the number of animals exhibiting detectable pulses was reduced in UR when compared to CR. At 12 days after OVX, 100% of CR showed detectable pulses at all doses of E<sub>2</sub>, while 100, 60, 50 and 17% of UR had detectable pulses at 0, 1.2, 2.4 and 4.8 µg doses of E<sub>2</sub>, respectively. UR with detectable pulses had mean LH concentrations which did not differ from those of CR, but UR which no longer exhibited detectable pulses had mean LH concentrations significantly lower than CR. These results suggest that the suppression of LH pulses by E<sub>2</sub> in UR may be an ‘all-or-none’ phenomenon, more likely to occur at a hypothalamic (or other CNS) site than at the pituitary level.