Acute administration of electroconvulsive shock (ECS) has been shown previously to produce potent analgesia which is only partially reversed by naloxone but shows almost complete tolerance after both repeated ECS and chronic morphine administration. In an attempt to elucidate the underlying basis of ECS analgesia it was recently compared with the analgesic effect of a newly identified opioid, humoral (H)-endorphin. Intracerebroventricular (i.c.v.) injection of H-endorphin to rats produces a dose-related analgesic effect as measured by the tail flick method. Furthermore, 4 days of daily i.c.v. injections of 40 microgram of morphine resulted in complete tolerance to the analgesic effect of H-endorphin. However, naloxone only caused a partial reversal of H-endorphin analgesia. Surprisingly only the lower dose of 1 mg/kg exerted a significant antagonistic effect while a higher dose of 10 mg/kg of the antagonist was without effect. A similar unconventional profile of the effect of naloxone could be seen with ECS analgesia. Here, pretreatment with 1 mg/kg of naloxone significantly attenuated analgesia, whereas administration of 10 mg/kg of the antagonist was without effect. In contrast, catalepsy measured in the same animals was not affected by 1 mg/kg of naloxone but increasing the dose to 10 mg/kg produced a significant attenuation of ECS catalepsy. The opioid nature of H-endorphin analgesia on one hand and the unconventional dose relation with naloxone of both H-endorphin and ECS analgesia on the other hand, suggests the involvement of this opioid in analgesia induced by ECS. Furthermore, it is possible that other behavioral manipulations which display only partial opiate characteristics may be mediated by H-endorphin or similar endogenous substances.