Myocardial ischaemia and left ventricular dysfunction have been described in cocaine users. Whether nitric oxide (NO) inhibition may potentiate the effects of cocaine on coronary circulation and ventricular function is still unknown. In order to test this hypothesis, 38 pentobarbital-anaesthetized pigs were instrumented for systolic blood pressure, coronary blood flow, left ventricular dp/dt, cardiac output, left ventricular end-diastolic and end-systolic lengths and shortening fraction. The pigs were randomized into three groups: control group: i.v. saline (n = 5); group 1: i.v. cocaine, 10 mg kg-1 over 20 min (n = 17); group 2: the same doses of cocaine 30 min after i.c. L-NAME 20 microg/kg min-1 infusion (n = 16). In order to know whether the observed effects were specific of NO inhibition, in five pigs i.c. L-arginine was simultaneously infused with L-NAME, in five pigs i.c. NTG, an endothelial-independent vasodilator, was simultaneously infused with L-NAME before cocaine was administered, and in nine additional pigs the proximal left anterior descending (LAD) flow was reduced to around 20% of the basal value by means of a mechanical occluder before cocaine was administered. Cocaine i.v did not change the coronary blood flow, while it induced a significant reduction in cardiac output, left ventricular dp/dt and shortening fraction (15 +/- 4-8 +/- 4%, P < 0.05). When cocaine was administered after L-NAME infused i.c. during 30 min, a significantly more severe reduction of the shortening fraction (12 +/- 3-4 +/- 2%, P < 0.0001) was induced; this effect was abolished by simultaneous perfusion of L-arginine i.c. NTG. The results when cocaine was administered after the 20% LAD flow reduction by mechanical occluder did not differ from those of cocaine alone. NO inhibition intensifies the cocaine-induced left ventricular dysfunction.