We previously reported that saturable stereospecific binding of [3H]-naltrexone in rat brain homogenates prepared from castrated male rats was greater than the corresponding binding in intact animals. We now report that we have replicated these results and that the difficulty of other investigators in observing these differences is due to methodological factors. Specifically, when samples were filtered individually and rapidly, differences between castrated and intact rats were maintained. The increase in binding was also observed when tissues were washed to remove endogenous opioids prior to incubation, when [3H]-naloxone was used as the ligand, and when various antagonists were used as displacers in the radioreceptor assay.