In the rat isolated vas deferens, methoxamine 8.1 microM produced an initial phasic response that declined towards baseline and was followed by rhythmic contractions that continued until wash-out. These responses were predominant in the epididymal half. BaCl2 1 mM produced a similar type of response which was not mediated by noradrenaline release or activation of alpha-adrenoceptors. The barium responses were similar in the epididymal and prostatic halves. Incubation in nominally Ca2+-free solution caused abolition or near abolition of rhythmic contractions produced by barium or methoxamine. The initial phasic response to methoxamine was abolished in Ca2+-free solution, whereas that produced by barium persisted. Rhythmic contractions produced by methoxamine or barium were inhibited by Mg2+ (2.4-20 mM) and by La3+ (1-5 mM). Mg2+ had selectivity for inhibition of the frequency of methoxamine- but not barium-induced rhythmic contractions. Despite their dependence on [Ca2+]o, barium- and methoxamine-induced rhythmic contractions were resistant to inhibition by calcium channel inhibitors. Verapamil, nifedipine and flunarazine inhibited the amplitude of rhythmic contractions more readily than the frequency (methoxamine IC50 for verapamil: amplitude = 29.8 +/- 5.40 microM, n = 6, frequency = 96.7 +/- 31.0 microM, n = 5, for nifedipine: amplitude = 2.42 +/- 0.34 microM, n = 7, frequency = 3.24 +/- 0.75 microM, n = 7, and for flunarizine: amplitude = 15.9 +/- 5.95 microM, n = 7, frequency = 153 +/- 28.6 microM, n = 7). There was no differentiation between inhibition of methoxamine and barium-induced responses. Like Mg2+, methoxyverapamil selectively inhibited the frequency of methoxamine-induced contractions (IC50: amplitude = 16.8 +/- 2.86 microM, n = 5, frequency = 2.07 +/- 0.81 microM, n = 5) but not barium-induced contractions (IC50: amplitude = 13.9 +/- 1.95 microM, n = 5, frequency = 48.5 +/- 8.98 microM, n = 5). Diazoxide (43.3-2167 microM) and nitroprusside (3.36-6712 microM) had only a small effect on barium contractions, but produced a dose-related reduction in the amplitude of methoxamine-induced responses. Diazoxide and nitroprusside caused methoxamine contractions to occur in groups, although they had no effect on their overall frequency. It is concluded that barium- and methoxamine-induced rhythmic contractions in the rat vas deferens are mediated by the entry of [Ca2+]o via membrane calcium channels that have a lower affinity (10-100 X) for calcium channel inhibitors than those mediating the KCl response. Channels activated by methoxamine are concentrated in the epididymal half, whereas those opened by barium are evenly distributed. However, although responses to methoxamine and barium are similar in form, differences in the effects of some of the drugs tested, together with the results of previous studies, indicate that they produce contractions by different mechanisms.