The hypothesis that mu-opioid agonists having low antinociceptive efficacy might be more susceptible to interference with G-protein coupling than mu-opioid agonists having higher antinocicep-tive efficacy was tested. Supraspinal antinociceptive efficacy for the three mu-opioid agonists morphine, [D-Ala(2), NMePhe(4), Gly(5)-ol]-enkephalin (DAMGO) and sufentanil in the mouse 55 degrees C warm-water tail-flick test was evaluated 18-24 h after intracerebroventricular (i.c.v.) administration of beta-funaltrexamine (beta-FNA). The beta-FNA pretreatment (0.2-2.0 nmol) attenuated antinociception in the order morphine > DAMGO > sufentanil, consistent with previous reports of their relative antinociceptive efficacy. The association of efficacy with G-protein coupling was then assessed by determining sensitivity to i.c.v. (0.1-3.0 mu g) pertussis toxin (PTX) or cholera toxin (CTX). The effect of PTX on equiantinociceptive doses was in the inverse order of agonist efficacy. CTX augmented sufentanil-induced antinociception. Morphine- and DAMGO-induced antinociception were unaffected by CTX. These data suggest that: (i) highly efficacious mu agonists (viz., sufentanil) couple more efficiently to PTX-sensitive inhibitory G(i)-proteins than do agonists of lower efficacy (viz., morphine, DAMGO) and (ii) highly efficacious mu agonists have greater capacity to utilize CTX-sensitive stimulatory G(s)-proteins than do mu-agonists with lower efficacy.