Apomorphine is a non-selective dopamine D1/D2 receptor agonist, which has been used for patients with Parkinson's disease and reported to induce QT interval prolongation and cardiac arrest. To clarify their causal link, we assessed the cardiovascular and pharmacokinetic profile of apomorphine with the halothane-anaesthetized canine model (n = 4), whereas pro-arrhythmic potential of apomorphine was analysed with the chronic atrioventricular block canine model (n = 4). In the halothane-anaesthetized model, 0.01 mg/kg, i.v. of apomorphine hydrochloride over 10 min., providing about 10 times of its therapeutic concentration, increased the heart rate and ventricular contraction; 0.1 mg/kg over 10 min., providing about 100 times of the therapeutic, prolonged the ventricular effective refractory period; and 1 mg/kg over 10 min., providing about 1000 times of the therapeutic, decreased the ventricular contraction, mean blood pressure and cardiac output together with the intraventricular conduction delay and prolongation of the effective refractory period, whereas the left ventricular end-diastolic pressure, atrioventricular nodal conduction or ventricular repolarization were hardly affected. Meanwhile, in the atrioventricular block model, 1 mg/kg, i.v. of apomorphine hydrochloride over 10 min. neither prolonged the QT interval nor induced torsade de pointes. These results suggest that apomorphine may possess a wide margin of cardiovascular safety contrary to our expectations.