1. In order to characterize P2-purinoceptor(s) in human urinary bladder the contractile effects of ATP and its slowly-hydrolyzable analogues alpha, beta-methylene ATP (alpha, beta-MeATP) and beta, gamma-methylene ATP (beta, gamma-MeATP) were investigated on human detrusor strips taken from patients undergoing cystectomy for bladder carcinoma. 2. Serial concentration-response curves (SCRC) for ATP, alpha, beta-MeATP and beta, gamma-MeATP were constructed with an interval of 25 min between two successive doses to avoid tachyphylaxis. ATP (10 microM-10 mM) induced a phasic contraction, which was very rapid in onset. The dose-response curve to ATP appeared not to be monophasic: at the lower concentrations (10-300 microM) the curve was shallow, whilst at high concentrations (1-10 mM) the curve was steeper. The magnitude of the response obtained at the highest concentration tested (10 mM) was only 21.1 +/- 2.8% (mean +/- s.e. mean; n = 4) of the KCl (100 mM)-induced contraction. 3. alpha, beta-MeATP (0.3 microM-1 mM) and beta, gamma-MeATP (10 microM-1 mM) elicited a phasic contraction with a time course similar to that exhibited by ATP. The magnitude of the response obtained at the highest concentration tested (1 mM) was 70.3 +/- 6.3% for alpha, beta-MeATP (n = 10) and 27.9 +/- 4.5% for beta, gamma-MeATP (n = 8) of KCl (100 mM)-induced contraction. The rank order of potency was alpha, beta-MeATP > beta, gamma-MeATP > ATP. A plateau of response could not be achieved by any of these agonists. 4. The P2-purinoceptor antagonist, suramin (10-300 microM), dose-dependently antagonized only the lower part of alpha,beta-MeATP dose-response curve. Data were analysed in terms of dose-ratio estimated at two levels of response (10% and 35% of KC1 100 mM-induced contraction). At 10% of KCl response the Schild plot slope was 0.98 and the estimated pKB was 5.85, whereas using the dose-ratio at the 35% level of the KCl response, the Schild plot was not linear suggesting an interaction of alpha,beta-MeATP with a heterogeneous receptor population.5. The putative P2-purinoceptor antagonist, Coomassie Brilliant Blue G (CB-G) at 0.3 and 1 l micro M(n = 5), shifted to the left the alpha,beta-MeATP SCRC. The response at the highest concentration of agonist was potentiated, being equal to 78.8 +/- 11.7% of the KCl (100 mM) response (n = 5). CB-G at 0.3 microM also shifted to the left the beta,upsilon-MeATP SCRC and significantly potentiated the response at 1 mM up to 46.3 +/- 5.6% of KCl 100 mM response (n = 4).6. Pretreatment with terodotoxin (TTX) at 1 microM shifted to the left the alpha,beta-MeATP SCRC but the response to the highest concentration of the agonist was not potentiated, being 73.6 +/- 9.9% of the KCl(100 mM) response (n = 5). TTX (1 micro M) shifted to the left the beta,upsilon-MeATP SCRC and significantly potentiated the response at 1 mM (61.6 +/- 3.1% of KCl response; n = 4).7. The NO synthase inhibitor NG-nitro-L-arginine methyl ester (L-NAME) at 100 micro M did not modify the SCRC to either alpha, beta or beta,upsilon-MeATP.8. We conclude that in human detrusor muscle there is a heterogeneity of purinoceptors. The complex antagonism exhibited by suramin suggests the presence not only of Ph-purinoceptors but also of another contractile P2-purinoceptor subtype insensitive to suramin. Moreover, the activity of CB-G and TTX seems to support the existence of a prejunctional P2-purinoceptor subtype inducing the release of one or more inhibitor neurotransmitters.