We tested the hypothesis that nitric oxide (NO) and adenosine contribute to the onset of vasodilation during dynamic forearm exercise. Twenty-two subjects performed rhythmic forearm exercise (20 % of maximum) during control and NO synthase (NOS) inhibition (N (G)-monomethyl-L-arginine; L-NMMA) trials. A subset of subjects performed a third trial of forearm exercise during combined inhibition of NOS and adenosine (aminophylline; n = 9). Additionally, a separate group of subjects (n = 7) performed rhythmic forearm exercise during control, inhibition of adenosine alone and combined inhibition of adenosine and NOS. Forearm vascular conductance (FVC; ml min(-1) · 100 mmHg(-1)) was calculated from blood flow and mean arterial pressure (mmHg). The onset of vasodilation was assessed by calculating the slope of the FVC response for every duty cycle between baseline and steady state, and the number of duty cycles (1-s contraction/2-s relaxation) to reach steady state. NOS inhibition blunted vasodilation at the onset of exercise (11.1 ± 0.8 vs. 8.5 ± 0.6 FVC units/duty cycle; P < 0.001 vs. control) and increased the time to reach steady state (25 ± 1 vs. 32 ± 1 duty cycles; P < 0.001 vs. control). Vasodilation was blunted further with combined inhibition of NOS and adenosine (7.5 ± 0.6 vs. 6.2 ± 0.8 FVC units/duty cycle; P < 0.05 vs. L-NMMA alone), but not with aminophylline alone (16.0 ± 2.2 vs. 14.7 ± 2.0 FVC units/duty cycle; P = 0.67 vs. control). Our data indicate that NO and adenosine (in the absence of NO) contribute to the onset of vasodilation during dynamic forearm exercise.