Twelve depressed patients were allocated to amitriptyline (7 days 75 mg nocte, increasing to 150 mg nocte) or zimelidine (100 mg nocte increasing to 200 mg nocte) in a randomized double-blind design. Cardiovascular assessment was made at pre-drug, 7 and 14 days using thoracic impedance cardiography, a non-invasive technique for measurement of both cardiac output (CO) and systolic time intervals. Neither drug significantly changed systolic time intervals. Amitriptyline 75 mg significantly reduced stroke volume (SV) (sitting P less than 0.02, standing P less than 0.01) with a similar trend at 150 mg. Zimelidine tended to increase SV. Heart rate (HR) was increased by amitriptyline 75 mg (P less than 0.02) but zimelidine induced no consistent changes in HR. Amitriptyline 75 mg in the standing data reduced CO (P less than 0.005). In contrast zimelidine 100 mg increased sitting CO (P less than 0.05) and this trend was seen at 200 mg in both sitting and standing data. The results suggest that zimelidine has low cardiotoxicity and may have a role in the treatment of depressed patients with cardiovascular disease.