In two experiments involving ovariectomized Scottish Blackface ewes and steroid replacement therapy, the effects of level of food intake and of body condition on hypothalamic and pituitary sensitivity to progesterone and oestradiol feedback were investigated. In experiment 1, groups of 16 ewes in similar levels of body condition were given dried grass pellets ad libitum (mean dry-matter (DM) intake of 2481 (s.e. 71) g per head per day; group H) or 1105 g DM per head per day (group M). In experiment 2, groups of 16 ewes were fed so that they achieved body condition scores of 2·70 (s.e. 0040; group HBC) or 1·67 (s.e. 0·036; group LBC) prior to the start of the experiment. All ewes were then fed to maintain their respective levels of body condition. In both experiments, all ewes were ovariectomized and intravaginal progesterone pessaries and subcutaneous oestradiol implants were inserted; these were designed to induce circulating concentrations of these steroids within the normal range for intact ewes. Approximately 1 week later, blood samples were collected at 15-min intervals for 10 h, and at 10-min intervals for 8 h before and after pessary withdrawal respectively. All samples were assayed for LH and selected samples were assayed for FSH and progesterone. Gonadotropin profiles were also recorded during 2-h periods following injection (i.v.) of 0·25 μg and 5 μg GnRH. When circulating progesterone concentrations were elevated, due to the presence of a progesterone pessary, there was no effect of level of food intake on mean LH or FSH concentrations or mean LH pulse frequency but H ewes had a higher mean LH pulse amplitude than M ewes (P < 005). When the progesterone pessaries were absent, H ewes had a higher mean LH concentration (P < 001), mean LH pulse frequency (P < 001) and LH pulse amplitude (P < 0·01) than M ewes but mean FSH concentrations were similar in the two groups. Irrespective of the presence or absence of elevated progesterone concentrations, the pituitary responses to GnRH injection were either not significantly different or were lower in H compared with M ewes. In experiment 2, HBC ewes had a higher mean LH concentration (P < 001), LH pulse frequency (P < 0·001), pulse amplitude (P < 0·001) and FSH concentration (P < 001) than LBC ewes, irrespective of circulating progesterone concentrations. The mean heights and ‘areas’ of the LH peak induced by injection of 0·25 μg GnRH were larger in HBC than LBC ewes (P < 005) irrespective of circulating progesterone concentrations. When mean concentrations before injection were taken into account, the corresponding increases in FSH concentration were similar for the two groups. Following injection of 5 u.g GnRH, the differences in peak size were generally not significant. It is concluded that during the follicular phase of the cycle, a low level of intake is associated with increased hypothalamic sensitivity, but not pituitary sensitivity to oestradiol feedback. Body condition affects both hypothalamic and pituitary activity but the relative importance of direct and indirect effects of body condition on the sensitivity of these organs to steroid feedback, remains to be elucidated.