Openers of K_ATP channels are known to inhibit KCl-, carbachol- and also electrically induced contractions in detrusor muscle strips from various species. Contractions of isolated strips of urinary bladder are usually of higher amplitude when the urothelium has been removed. This has been explained by the release of an urothelium-derived relaxing factor. In this study we examined whether intact urothelium may modulate the effect of the selective K_ATP channel opener rilmakalim. Contractile responses to 85 mM KCl and 10 μM carbachol were measured in detrusor strips from mouse, pig and man. In the presence of an intact urothelium, contractions were significantly reduced in strips from all three species investigated. In preparations with urothelium rilmakalim reduced KCl contractions with similar potency and efficacy [-logIC₅₀ (M) 4.6 to 5.1; Eff_max reduction to 14-30% of control]. However, in urothelium-denuded strips rilmakalim was more potent in pig (-logIC₅₀ 5.5) than in mouse and man (-logIC₅₀ 4.7 and 4.4, respectively). The order of potency for rilmakalim to suppress carbachol-induced contractions was pig (-logIC₅₀ 6.7) > man (5.8) > mouse (4.7); contractions were significantly more reduced in pig (Eff _max reduction to 11±2%, n=10) and in mouse (21±2%, n=8) than in human detrusor (55±5%, n=5). The presence of urothelium did not affect the concentration-response curves for rilmakalim, with the exception of KCl-induced contractions in pig. Only the rilmakalim-induced relaxation of carbachol-mediated contractions in pig were prevented by the K_ATP channel blocker glibenclamide. We conclude that with this one exception, the responses to rilmakalim in detrusor contractions were not mediated by K _ATP channel opening.