We previously demonstrated a bimodal distribution of vasodilator responsiveness to adenosine (Ado) infusion in human subjects, despite similar responses to exercise between subgroups [subjects responsive to Ado infusion (Ado responders) and subjects with blunted vasodilator responses to Ado infusion (Ado nonresponders]). (Martin EA, Nicholson WT, Eisenach JH, Charkoudian N, and Joyner MJ. J Appl Physiol 101: 492-499, 2006). A component of this difference was attributed to a larger nitric oxide component of Ado-mediated vasodilation in responders. However, there may also be differences in Ado receptors between these subgroups. We hypothesized that Ado receptor antagonism would reduce vasodilator responsiveness to Ado and exercise only in Ado responders. To test this hypothesis, we compared forearm vasodilation induced by intra-arterial infusion of three doses of Ado to vasodilation during three workloads of forearm handgrip exercise before and after Ado receptor antagonism with aminophylline (Aph) in 19 subjects. In Ado responders, the change in forearm vascular conductance above baseline for the low, medium, and high doses of Ado, respectively, was 93 +/- 16, 140 +/- 14, 194 +/- 18 before Aph and 27 +/- 12, 71 +/- 19, and 134 +/- 34 ml.min(-1).100 mmHg(-1) after Aph (P < 0.05 for low and medium dose before vs. after Aph). For nonresponders, these values were 30 +/- 5, 39 +/- 6, and 78 +/- 9 ml.min(-1).100 mmHg(-1) before Aph (P < 0.05 vs. responders), with no difference after Aph (P > 0.05). We found that Ado receptor blockade significantly inhibited exercise hyperemia only at high workloads in both responders and nonresponders (P < 0.05 before vs. after Aph). We conclude that there may be reduced Ado receptor responsiveness or sensitivity in nonresponders. Furthermore, Ado may play a limited role exercise hyperemia in both subgroups.