To test the ability of ketone bodies to inhibit myocardial fatty acid oxidation in vivo, the myocardial clearance kinetics of [1-11C]palmitate was assessed with positron emission tomography in six fasted volunteers and six instrumented dogs, studied repeatedly before and during infusion of 3-hydroxybutyrate (17 mumol.kg-1 x min-1). With the use of multiexponential fitting of tissue time-activity curves, the size, half time (T1/2), and index of the early rapid phase of 11C myocardial clearance, reflecting palmitate oxidation, were calculated. In humans, the relative size (-28%, P < 0.001) and index (-37%, P < 0.01) of the early rapid phase decreased significantly during infusion of 3-hydroxybutyrate, consistent with decreased fatty acid oxidation. Paradoxically, T1/2 decreased from 10.1 +/- 1.6 to 7.4 +/- 1.1 min (P < 0.01). To elucidate possible mechanisms, multiple coronary arteriovenous samples were obtained from the dogs to assess the efflux of oxidized and nonmetabolized tracer. Infusion of 3-hydroxybutyrate resulted in decreased myocardial [11C]CO2 production (-40%, P < 0.05) and reduced palmitate retention (-38%, P < 0.05). In three dogs, the arteriovenous difference in radiolabeled palmitate became negative 10 min after injection, indicating backdiffusion of nonmetabolized tracer from the myocardium. Thus a steady-state infusion of 3-hydroxybutyrate, resulting in physiological plasma levels, alters [1-11C]palmitate kinetics in vivo by decreasing myocardial long-chain fatty acid oxidation and by increasing backdiffusion of nonmetabolized tracer.