The possibility that isosorbide dinitrate (ISDN) inhibits platelet function in humans has been explored in vitro and in vivo. Incubation of citrated platelet-rich plasma from healthy subjects with scalar concentrations (1.25, 12.5 and 125 micrograms/ml) of ISDN for 5 and 10 minutes resulted in a decrease in platelet aggregation after ADP, adrenaline, and arachidonic acid at the highest drug concentration (mean decrease: 72% [p less than 0.01], 56% [p less than 0.05] and 62% [p less than 0.05], respectively, with the 10-minute incubation). Also, a significant reduction (30%) in generated thromboxane (TX)B2 levels was observed after arachidonic acid (p less than 0.01). ISDN was then infused at rate of 4 mg/hour for 30 minutes in 11 patients with angina and at a rate of 30 mg/hour for 20 minutes in 8. The smaller dose, which caused minor changes in arterial pressure and heart rate, was accompanied by a marked, significant decrease in ADP- and adrenaline-induced aggregation, with a nadir at 60 minutes from the infusion stop (decreases of 40% and 51% respectively). Circulating platelet aggregates also decreased, with a minimum (-41%, p less than 0.05) at the end of the infusion. The higher infusion rate, causing marked hemodynamic effects, was not accompanied by the occurrence of clear antiplatelet effects. Thus, ISDN can affect platelet function both in vitro and in vivo. The in vivo effect occurs at lower concentrations than in vitro but is blunted when a marked hemodynamic response occurs.