Cumulative concentration-effect curves for prostaglandin E(2), sulprostone and butaprost were constructed in matched strips of human non-pregnant myometrium from 14 different donors. All samples were obtained from the mid-lateral wall of the uterus. Prostaglandin E(2) produced four types of concentration-effect curves: monophasic inhibitory (n = 7), monophasic excitatory (n = 2), biphasic consisting of an excitatory phase followed by an inhibitory phase (n = 4), and biphasic consisting of an inhibitory phase followed by an excitatory phase (n = 1). Sulprostone produced excitation of spontaneous contractile activity in all tissues (mean pEC(50) = 9.1+/-0.2, range 8.1-10.1, n = 14). Butaprost produced relaxation of cloprostenol-stimulated contractile activity in all tissues (mean pEC(50) = 5.7 +/- 0.1, range 5.0-6.9, n = 14). The mean pEC(50) value for sulprostone was significantly higher in tissues where prostaglandin E(2) caused some excitation (pEC(50) = 9.4 +/- 0.2, n = 7) compared to those where prostaglandin E(2) caused only inhibition (pEC(50) = 8.8 +/- 0.2, n = 7). Mean pEC(50) values for butaprost were not significantly different between these groups. These data suggest that (a) variability in EP receptor-mediated responses exists within a single anatomical site; (b) both excitatory and inhibitory EP receptor-mediated pathways are always operative in human non-pregnant myometrium, regardless of the type of tissue response to prostaglandin E(2); and (c) regulation of EP receptor-mediated responses occurs predominantly in the excitatory (EP(3) or EP(1) receptor) pathway rather than the inhibitory (EP(2) receptor) pathway.