A pipe for administration of inhaled cocaine and its pyrolytic products in laboratory animals was developed and tested. In-vitro trials showed 30.0 +/- 5.2% (mean +/- SE) recovery of cocaine in solvent. Five non-pregnant ewes were instrumented with tracheal T-tubes and vascular catheters. After surgical recovery, ewes received three doses of cocaine (free base) in a randomized fashion; 2 mg/kg and 4 mg/kg both by inhalation, and 2 mg/kg intravenously. Arterial blood samples were collected and assayed for cocaine and its major metabolites by high performance liquid chromatography. Blood pressure and heart rate were continuously recorded. Cocaine administered by inhalation was eliminated with a half-life of 1.6 +/- 0.5 min (mean +/- SE) compared to 3.4 +/- 0.9 following intravenous administration (p less than 0.03). Likewise, clearance values were greater following inhalation, 5532 +/- 1756 ml/min/kg, than following intravenous administration, 163 +/- 20.6 ml/min/kg (p less than 0.04). Both routes of administration led to significant elevations in blood pressure, 7.5% increase after smoking vs 20% increase after intravenous administration. No correlation was found between inhalational dose of cocaine and peak plasma cocaine concentration.