The effects of morphine on the secretion of luteinizing hormone (LH) were examined in male rats at intervals after castration. We found that morphine was extremely effective in suppressing serum LH levels in animals that had been castrated for periods of less than 7 days, but was considerably less potent in long-term castrates (13 + days). For example, the dose of morphine producing the half-maximal suppression of serum LH levels in 3-day castrates was 1.5 mg/kg, whereas in 31-day castrates the ED50 was 13.8 mg/kg. This insensitivity to morphine satisfied the two pharmacological criteria for tolerance: a parallel shift to the right in the morphine dose-response curve and a reduced effect of the drug at the same brain concentration. Hence, castration appeared to render male rats, never exposed to opiates, tolerant to the effects of morphine. Corresponding to the development of tolerance to morphine, 31-day castrates were also less responsive to the LH-depressing effects of testosterone than were 3-day castrates or sham-operated controls. In marked contrast to these results, castration did not significantly affect naloxone-induced increases in serum LH levels. The tolerance observed to morphine in the long-term castrated rat was selective to LH secretion because long- and short-term castrates and sham-operated controls were equally responsive to the antinociceptive effects of morphine (in fact, at a dose of 8 mg/kg long-term castrates were more sensitive to morphine than short-term castrates or sham-operated controls) and there was no apparent shift in the LD50. The mechanisms underlying the development of tolerance to morphine in castrated male rats are not clear at the present time.