1. Forearm blood flow responses to incremental challenges of acetylcholine and substance P, administered via the brachial artery, were measured by venous occlusion plethysmography in eight subjects in the presence of saline, the nitric oxide synthase inhibitor, NG-monomethyl-L-arginine, and a control vasoconstrictor, noradrenaline. 2. Substance P and acetylcholine caused dosedependent increases in forearm blood flow (P < 0.001). When separated by 30 min saline infusions, repeated responses did not undergo tachyphylaxis. 3. Noradrenaline caused a mean reduction in basal blood flow of 34-51% (P < 0.001), and augmented the percentage increases in blood flow with both substance P (P = 0.05) and acetylcholine (P = 0.03) infusions. 4. NG-Monomethyl-L-arginine caused a mean reduction in basal blood flow of 42-45% (P < 0.001) and significantly inhibited the responses to both substance P (P < 0.001) and acetylcholine (P = 0.05). 5. In comparison with saline responses, NG-monomethyl-L-arginine caused a mean inhibition of 69 +/- 8% for substance P-induced vasodilatation and 40 +/- 5% for acetylcholine-induced vasodilatation. However, comparing responses with those to the control vasoconstrictor noradrenaline, NG-monomethyl-L-arginine caused a mean inhibition of 81 +/- 5% for substance P responses and 58 +/- 3% for acetylcholine responses. Inhibition by NG-monomethyl-L-arginine of the response to substance P was significantly greater than inhibition of the response to acetylcholine (P = 0.02). 6. Hence, in healthy men, a greater proportion of the forearm vasodilatation to substance P than to acetylcholine appears to be nitric oxide-mediated. Given its greater stability, substance P may be more suitable as a pharmacological tool in the investigation of stimulated nitric oxide production and endothelial cell function.