This study was essentially an in vivo protection experiment designed to test further the hypothesis that stress induces release of endogenous opioids which then act at opioid receptors. Rats that were either subjected to restraint stress for 1 hr or unstressed were injected ICV with either saline or 2.5 micrograms of beta-funaltrexamine (beta-FNA), an irreversible opioid antagonist that alkylates the mu-opioid receptor. Twenty-four hours later, subjects were tested unstressed for morphine analgesia (tail-flick assay) or were sacrificed and opioid binding in brain was determined. [3H]D-Ala2NMePhe4-Gly5(ol)enkephalin (DAGO) served as a specific ligand for mu- opioid receptors, and [3H]-bremazocine as a general ligand for all opioid receptors. Rats injected with saline while stressed were significantly less sensitive to the analgesic action of morphine 24 hr later than were their unstressed counterparts. Beta-FNA pretreatment attenuated morphine analgesia in an insurmountable manner. Animals pretreated with beta-FNA while stressed were significantly more sensitive to the analgesic effect of morphine than were animals that received beta-FNA while unstressed, consistent with the hypothesis that stress induces release of endogenous opioids that would protect opioid receptors from alkylation by beta-FNA. beta-FNA caused small and similar decreases in [3H]-DAGO binding in brain of both stressed and unstressed animals. Stressed rats injected with saline tended to have increased levels of [3H]DAGO and [3H]-bremazocine binding compared to the other groups. This outcome may be relevant to the tolerance to morphine analgesia caused by stress.