1. The aim of this study was to determine whether the site of action of relaxin as a relaxant of rat myometrium is at the cell membrane or at an intracellular-site. Therefore, the potency of relaxin was determined against spasms reliant predominantly upon either extracellular Ca2+ or intracellular Ca2+. Uterine spasms dependent upon extracellular Ca2+ were elicited by (i) oxytocin (0.2 nM) (ii) Bay K 8644 (1 microM) in 10 mM K(+)-rich PSS and (iii) KCl (80 mM). Uterine spasm dependent upon intracellular Ca2+ was elicited by oxytocin (20 nM) in the presence of nifedipine (500 nM). The effects of relaxin against these spasmogens were compared with those of levcromakalim, nifedipine and salbutamol. 2. Relaxin (0.2-6.3 nM), levcromakalim (25-800 nM), salbutamol (1-63 nM) and nifedipine (1-250 nM) caused concentration-dependent inhibition of the spasm evoked by oxytocin (0.2 nM) and relaxin was the most potent relaxant. 3. Relaxin and nifedipine were slightly less potent against the spasm induced by Bay K 8644 (1 microM) than against spasm induced by oxytocin (0.2 nM) (15 fold and 13 fold respectively). Levcromakalim and salbutamol were equipotent against the spasm evoked by Bay K 8644 (1 microM) and that evoked by oxytocin (0.2 nM). 4. Relaxin induced only 47 +/- 7% inhibition of the KCl (80 mM)-evoked spasm at a concentration of 0.8 microM. Levcromakalim was much less potent (427 fold) against the spasm evoked by KCl (80 mM) than against the spasm evoked by oxytocin (0.2 nM). The potency of salbutamol against the spasm evoked by KCl (80 mM) was modestly reduced (14 fold) compared to that against the spasm evoked by oxytocin (0.2 nM). The potency of nifedipine against the KCl (80 mM)-evoked spasm was not different from that against the oxytocin (0.2 nM)-evoked spasm. 5. The potencies of relaxin and levcromakalim against the spasm evoked by oxytocin (20 nM) + nifedipine (500 nM) were greatly reduced (74 fold and 234 fold respectively) compared to their potencies against the spasm evoked by oxytocin (0.2 nM). The potency of salbutamol against these two spasmogens was not different. 6. Relaxin was much less potent against the spasm dependent upon intracellular Ca2+ (that induced by oxytocin (20 nM) + nifedipine (500 nM)) than against the spasms dependent upon extracellular Ca2+, those induced by oxytocin (0.2 nM) and Bay K 8644 (1 microM). In this regard, relaxin resembled levcromakalim and nifedipine rather than salbutamol. Therefore, the major site of action of relaxin appears to be located at the plasma membrane rather than at an intracellular level. The observation that relaxin was less effective against the KCl (80 mM)-induced spasm than against the oxytocin (0.2 nM)-evoked spasm may indicate that relaxin has a minor action involving K(+)-channel opening. 7. High concentrations of relaxin (up to 1 microM) induced significant inhibition of the spasm dependent upon intracellular Ca2+. Thus at high concentrations relaxin also appears to have an additional intracellular action.