The effect of the nitric oxide synthase inhibitor N-G-Nitro-L-arginine methyl ester (L-NAME) towards muscarinic receptors was studied in vitro and in vivo. L-NAME displaced the antimuscarinic ligand [H-3]quinuclidinyl benzilate ([H-3]QNB) from its specific binding sites in rat cerebral cortex and cerebellum homogenates with a more than 10,000 fold lower affinity than atropine, pirenzepine and AFDX 116. Data for L-NAME binding were best fit according to a two-site model (K-d 7.2 nM and 3,000 nM) in the rat cerebellum, whereas in rat cortex a one-site model (K-d 1670 nM) was superior. In anesthetized rats and rabbits L-NAME (7.5-185 mu mol/kg) attenuated a hypotensive response to Acetyl beta-methyl-choline (Ac beta-Me Ch)(6.25 nmol/kg) in a dose related fashion, but this effect was negligible as compared to that of atropine (8.8 and 17.7 nmol/kg). Furthermore, the effect of L-NAME was not specifically antimuscarinic since its attenuating effect against ATP- or histamine-induced responses was not statistically different from that of Ac beta-Me Ch. A vagus stimulation induced bradycardia was entirely uninfluenced by L-NAME (37 mu mol/kg). In isolated bladder experiments (rabbit) we demonstrated a complete lack of efficacy of L-NAME against Ac beta-Me Ch induced contractions. In the pithed rat preparation L-NAME was ineffective against the McN-A-343 induced presser responses. In summary, we demonstrated that the nitric oxide synthase inhibitor L-NAME shows very weak affinity at M-1- and M-2-receptors in the rat brain in vitro, but appears to have no significant antimuscarinic properties against M-1-, M-2- and M-3-receptor mediated effects in rats and rabbits in vivo.