Eight time-dated pregnant ewes at 125 days' gestation (145 days = term) underwent surgery for placement of fetal vascular catheters, electrodes for recording fetal behavioural state, and maternal venous catheters. Three days later, fetal cerebral and myocardial blood flow were determined by the coloured microsphere technique under four conditions: (1) during rapid-eye movement (REM) sleep, before fetal cocaine infusion, (2) 30 min after initiation of a cocaine infusion to the fetus at 0.2 mg/kg per min, (3) during REM sleep, before maternal cocaine infusion, and (4) 30 min after initiation of a cocaine infusion to the ewe at 0.3 mg/kg per min. Cocaine infusion directly to the fetal lamb did not cause hypoxaemia or significantly change cerebral or myocardial blood flow or oxygen delivery. Cocaine administered to the ewe led to a drop in fetal oxygen tension from 3.0 +/- 0.5 to 2.5 +/- 0.3 kPa (P < 0.0001) and in fetal oxygen content from 3.8 +/- 0.7 to 2.8 +/- 0.4 mmol O-2/L (P < 0.0001). Prior to maternal cocaine administration, fetal cerebral blood flow was 146 +/- 103 mL/100 g per min and during maternal cocaine infusion it went to 184 +/- 147 mL/100 g per min (P = NS) while myocardial blood flow increased from 156 +/- 92 to 333 +/- 178 mL/100 g per min (P < 0.002). This increase in blood flow negated the effects of hypoxaemia so that cerebral oxygen delivery was unaffected while myocardial oxygen delivery increased an average of 67%. It is concluded that cocaine administration to pregnant sheep does not impede fetal cerebral or myocardial oxygen delivery.