The pronounced decrease in arterial blood pressure evident during anesthetic induction with propofol has raised the possibility that propofol has a direct negative inotropic effect. Previous attempts to evaluate this mechanism in vivo have been inconclusive because of confounding variables associated with intravenous administration of propofol. Accordingly, in the current study, steady-state changes in myocardial contractility and related hemodynamic parameters were assessed during intracoronary infusions of propofol in seven open-chest dogs anesthetized with fentanyl and midazolam. The left anterior descending coronary artery (LAD) was cannulated and perfused at controlled pressure (100 mmHg) with normal arterial blood. In LAD-perfused myocardium, contractility was evaluated from measurements of percent segmental shortening (%SS) obtained with ultrasonic crystals. Coronary blood flow in LAD was measured electromagnetically and used to calculate myocardial oxygen consumption (MVO2; Fick principle) and coronary propofol concentration. Propofol was infused into the LAD at 150, 300, 600, and 1,200 micrograms/min (P-150, P-300, P-600, P-1,200). These infusion rates yielded calculated blood concentrations of 7 +/- 1, 15 +/- 1, 26 +/- 2, and 50 +/- 5 micrograms.ml-1, respectively. The calculated blood concentrations at P-150 were in the clinical range, whereas those at P-300, P-600, and P-1,200 were supratherapeutic. P-150 had no effect on %SS, whereas higher infusion rates caused decreases in %SS. Changes in MVO2 by propofol generally paralleled changes in %SS. At P-150 and P-300, coronary blood flow was proportional to MVO2, whereas at P-600 and P-1,200, coronary blood flow was in excess of the prevailing MVO2, resulting in increased coronary venous oxygen tension.(ABSTRACT TRUNCATED AT 250 WORDS)