The ability of the imidazole derivative, ketoconazole, to inhibit thromboxane (Tx)A2 synthesis in response to ischemia was tested in ten volunteers. Two hours after taking placebo or ketoconazole 400 mg by mouth, plasma levels of the stable degradation product of TxA2, TxB2, were 300 +/- 129 pg/ml (mean +/- SEM) and 297 +/- 80 pg/ml, respectively. Arm ischemia for 10 min induced by inflation of a cuff to 220 mm Hg led to a rise in TxB2 levels to 657 +/- 157 pg/ml after placebo (p less than 0.05) and 337 +/- 81 pg/ml after ketoconazole. One hour after cuff deflation, TxB2 returned to pre-ischemia levels in both groups. Platelet TxB2 concentrations were 27 +/- 6 ng in the placebo and 35 +/- 6 ng/10(9) platelets in the ketoconazole group, and were unchanged by cuff inflation. The fact that plasma and platelet TxB2 values were not lower 2 hr after ketoconazole treatment was explored in another group of four nonstressed volunteers who received 400 mg of drug. After 2 hr, TxB2 values had fallen from 170 +/- 30 pg to 120 +/- 10 pg; at 4 hr, 6 hr, and 8 hr they were 30 +/- 20 pg, 5 +/- 5 pg, and 5 +/- 5 pg/ml, respectively. These results indicate that tourniquet ischemia provokes TxA2 synthesis, and that the source of this prostanoid is likely to be ischemic tissue and not platelets. Finally, ketoconazole can profoundly inhibit both background and stimulated TxA2 synthesis.