The increased tolerance to myocardial ischemia observed during the second of two sequential exercise tests, i.e. the warm-up phenomenon, has been proposed as a clinical model of ischemic preconditioning. Adenosine appears to be a mediator of ischemic preconditioning in both experimental and clinical settings. The purpose of this study was to investigate the role of A1 adenosine receptors in the warm-up phenomenon. A double-blind, placebo-controlled, cross-over design was used. Twelve patients with coronary artery disease and positive exercise test were randomized to receive either bamiphylline, a selective A1 adenosine receptor antagonist, or placebo, immediately prior to two consecutive treadmill exercise tests carried out on day 1. Then, on day 2 all patients underwent two consecutive exercise tests immediately after administration of the remaining treatment. During the first exercise test, bamiphylline, compared to placebo, increased the time to and rate-pressure product at 1.5 mm ST-segment depression (from 317 +/- 118 to 423 +/- 127 s, p < 0.05 and from 199 +/- 38 to 230 +/- 36 b/min.mmHg.10(2), p < 0.05, respectively). After both placebo and bamiphylline infusions, time to 1.5 mm ST-segment depression during the second exercise test was greater than that during the first test (445 +/- 121 vs 317 +/- 118 s, p < 0.001 and 483 +/- 128 vs 423 +/- 127 s, p < 0.05, respectively), as was rate-pressure product at 1.5 mm ST-segment depression (228 +/- 40 vs 199 +/- 38 b/min.mmHg.10(2), p < 0.01 and 253 +/- 42 vs 230 +/- 36 b/min.mmHg.10(2), p < 0.05, respectively). In conclusion, bamiphylline, at a dose able to increase ischemic threshold and exercise tolerance compared to placebo, does not prevent the warm-up phenomenon. These findings suggest that, in the setting of the warm-up phenomenon, A1 adenosine receptor blockade is insufficient to prevent ischemic preconditioning.