Abstinence signs were precipitated in rats by naloxone (1 mg - kg-1 s.c.) injected at various times (from 1.5 to 16 h) after a single dose of morphine hydrochloride (15 or 50 mg - kg-1 s.c.) administered in aqueous solution. Increasing the dose of morphine increased the latency of the phenomena and the duration of the underlying state; "shifts" of signs as described by Bläsig et al. (1974) in chronically morphinized rats also occurred when increasing the dose of morphine and the time interval between the injections of morphine and of naloxone. Naltrexone and diprenorphine were also effective. These three antagonists, given before morphine, were able to prevent precipitated abstinence: however, naloxone was almost ineffective when the higher dose of morphine was used and when the time interval was long. In these latter conditions, naltrexone was definitely more effective and longer acting and diprenorphine still more so. The same characteristics were found for the protective action of the three antagonists in acutely morphinized mice and the same order for their potencies in precipitating abstinence in acutely morphinized mice. Like naloxone, naltrexone and diprenorphine facilitated a nociceptive reaction in normal mice. The abstinence signs precipitated in acutely morphinized rats or mice are probably not unmasked excitatory effects of morphine as such effects should have been increased rather than inhibited by previous administration of specific antagonists; they might correspond to potentiated effects of the antagonists themselves. The prevention by specific antagonists of the abstinence syndrome is most simply interpreted by antagonism (direct or indirect) of dependence induction, but other interpretations are not excluded.