Plasma luteinizing hormone (LH) levels and LH responses to intravenous administration of 100 mug luteinizing hormone releasing hormone (LH-RH) were studied in entire rams, long-term castrated animals (operation performed six months previously), long-term castrated animals treated with testosterone for the two prededing weeks and short-term castrated animals (castrated 3 h before LH-RH injection). LH was measured by radioimmunoassay in samples taken at 5 or 15 min intervals. Basal LH levels were lower in entire rams (0-9 ng/ml) than in long-term castrated animals (6-0 ng/nl). After LH-RH treatment the LH response was much smaller (peak level 9-6 ng/ml), total response 13-3 ng/ml/1 h) and slower (120 min to peak) in entire than in long-term castrated animals (peak level 61-8 ng/ml, total response 141-2 ng/ml/1 h, 29 min to peak). Testosterone treatment after long-term castration depressed the basal LH level and delayed the peak LH response after LH-RH to values similar to those for entire rams. After short-term castration the response to LH-RH was already as great (peak level 70-1 ng/ml, total response 133-6 ng/ml/1 h) as after long-term castration. The latency to peak LH level (82 min) was intermediate between that for untreated and testosterone-treated long-term castrated animals (130 min). Testosterone treatment was considered to have acted on the hypothalamus to depress basal levels. The results provided evidence for the presence of two inhibitory actions of the testis at the pituitary level in the ram: a qualitative delaying action of testosterone and a quantitative inhibitory action of the testis on LH release after LH-RH injection. The latter may also be related to plasma testosterone levels.