W e assessed the effect of alcohol, before and after autonomic blockade, on left ventricular (LV) performance in conscious dogs. 10 animals were instrumented to determine LV volume from ultrasonic LV internal dimensions and measure LV pressure with a micromanometer. The animals were studied in the conscious state after full recovery from the operation. Blood alcohol was undetectable before and 67 +/- 14 mg/dl (mean +/- SD) at 20 min after alcohol administration. In response to alcohol, the LV systolic pressure was reduced slightly, the left ventricular end-diastolic pressure increased slightly. The maximum time derivative of LV pressure (dP/dtmax) and stroke volume were decreased. The end-systolic volume (VES), as well as effective arterial elastance, were significantly increased. There was no significant change in heart rate. Variably loaded pressure-volume loops were generated by acute caval occlusion before, immediately, and 20 min after the intravenous infusion of alcohol (0.2 g/kg). Three measures of LV performance were derived from these variably loaded pressure-volume loops: the end-systolic pressure-volume relation; the stroke work-end-diastolic volume relation; and maximum dP/dt-VED relation. The slopes of all three relations were significantly decreased in response to alcohol, and all three relations were shifted toward the right, indicating a depression of LV contractile performance. Similar, but greater depressions of LV performance with alcohol were observed following autonomic blockade. LV performance was restored by infusing dobutamine. We conclude that mildly intoxicating levels of alcohol (blood concentration less than 100 mg/dl) are capable of producing LV contractile depression in conscious animals, which is more marked after autonomic blockade. This suggests that patients with impaired LV function should avoid even small amounts of alcohol.