1 We have used the isolated, buer-perfused, superior mesenteric arterial bed of male and female rats to assess the relative contributions of nitric oxide (NO) and the endothelium-derived hyperpolarizing factor (EDHF) to endothelium-dependent relaxations to carbachol. 2 Carbachol caused dose-related relaxations of methoxamine-induced tone in mesenteric vascular beds from male rats described by an ED 50(M) of 0.43+0.15 nmol and a maximum relaxation (R max(M ) of 89.6+1.2% (n=28) which were not signi®cantly dierent from those observed in mesenteries from female rats (ED 50(F) =0.72+0.19 nmol and R max(F) =90.7+0.9%; n=22). 3 In the males, the addition of 100 mM N G -nitro-L-arginine methyl ester (L-NAME) caused the doseresponse curve to carbachol to be signi®cantly (P50.001) shifted to the right 15 fold (ED 50(M) =6.45+3.53 nmol) and signi®cantly (P50.01) reduced R max(M) (79.7+2.8%, n=13). By contrast, L-NAME had no eect on vasorelaxation to carbachol in mesenteries from female rats (ED 50(F) =0.89+0.19 nmol, R max(F) =86.9+2.3%, n=9). 4 Raising tone with 60 mM KCl signi®cantly reduced the maximum relaxation to carbachol in mesenteries from male rats 2 fold (R max(M) =40.3+9.2%, n=4; P50.001) and female rats by 1.5 fold (R max(F) =55.3+3.3%, n=6; P50.001), compared with methoxamine-induced tone. The potency of carbachol was also signi®cantly reduced 1.2 fold in preparations from males (ED 50(M) =0.87+0.26 nmol; P50.01) but not the females (ED 50(F) =4.04+1.46 nmol). In the presence of both 60 mM KCl and L-NAME, the vasorelaxation to carbachol was completely abolished in mesenteries from both groups. 5 The cannabinoid receptor antagonist SR141716A (1 mM), which is also a putative EDHF antagonist, had no signi®cant eect on the responses to carbachol in mesenteries from males or females (ED 50(M) =1.41+0.74 nmol, R max(M) =89.4+2.5%, n=7; ED 50(F) =2.17+0.95 nmol, R max(F) =89.9+1.8%, n=9). In mesenteries from male rats, in the presence of 100 mM L-NAME, SR141716A signi®cantly (P50.05) shifted the dose-response curve to carbachol 8 fold further to the right than that seen in the presence of L-NAME alone (ED 50(M) =53.8+36.8 nmol) without aecting R max(M) (72.4+4.8%, n=10). In mesenteries from female rats, the combined presence of L-NAME and SR141716A, signi®cantly (P50.01) shifted the dose-response curve to carbachol 7.5 fold, (ED 50(F) =6.66+2.46 nmol), as compared to L-NAME alone and signi®cantly (P50.001) decreased R max(F) (70.1+5.5%, n=8). 6 Vasorelaxations to the nitric oxide donor sodium nitroprusside (SNP), to the endogenous cannabinoid, anandamide (a putative EDHF) and to the ATP-sensitive potassium channel activator, levcromakalim, did not dier signi®cantly between male and female mesenteric vascular beds. 7 The continuous presence of sodium nitroprusside (SNP; 20 ± 60 nM) had no eect on vasorelaxation to carbachol in mesenteries from either males or females. In the presence of L-NAME, SNP signi®cantly (P50.05) reduced the potency of carbachol 6 fold, without aecting the maximal relaxation in mesenteries fr...