The ventricular tachycardia (VT) caused by high-dose tricyclic antidepressants has been hypothesized to be due to a quinidinelike effect with generation of repolarization abnormalities and afterdepolarizations. To test this hypothesis further, we infused amitriptyline in a graded fashion (0.5-1 mg/kg/min) in 23 chloralose-anesthetized dogs during endocardial monophasic action potential (AP) recording and continuous hemodynamic monitoring. Three groups of dogs were studied: group A (n = 5), crushed sinus node and fixed atrial pacing at 100 beats/min; group B (n = 12), crushed sinus node and fixed atrial pace plus intermittent accelerated pacing to mimic group C; and group C (n = 6) intact sinus node and unimpeded sinus tachycardia. Amitriptyline infusion induced VT in no (0 of 5) group A dogs, all (12 of 12) group B dogs during accelerated pacing, and 83% (5 of 6) of group C dogs. Dogs with VT had significantly higher heart rates (HR 184.8 +/- 39.3 beats/min) as compared with dogs without VT (115.2 +/- 12.5 beats/min, p = 0.0015). There was a strong positive correlation between the last RR coupling interval to the first VT interval (r = 0.85; p = 0.0033). Amitriptyline infusion caused rate-dependent QRS prolongation in each group, especially group C (p < 0.001). Action potential duration at 50% and 90% of repolarization (APD50, APD90) showed a biphasic response with progressive shortening followed by prolongation as amitriptyline serum concentrations increased. Afterdepolarizations were not detected from any monophasic AP recording, even in dogs with VT.(ABSTRACT TRUNCATED AT 250 WORDS)