Tricyclic antidepressants (TCAs) and selective serotonin reuptake inhibitors (SSRIs) are used to treat depression. Whereas cardiovascular effects have occasionally been reported during controlled studies with SSRIs, TCA treatment poses a well-known problem in this respect. To investigate the putative correlation between antidepressant dose or serum levels and adverse effects, the authors devised a naturalistic study to evaluate the tricyclic antidepressants' and SSRIs' effect on the cardiovascular system. The authors also compared antidepressant serum levels to adverse effects. Inpatients treated with TCAs or SSRIs were included; an electrocardiogram (ECG) and a Schellong test were carried out on the day patients entered the hospital and during steady-state treatment with antidepressant drugs when blood was drawn for therapeutic drug monitoring. The patient population consisted of 114 acutely depressed patients; 81 patients were treated with TCAs and 33 with SSRIs. The TCAs comprised amitriptyline (n = 43), clomipramine (n = 11), doxepin (n = 19) and imipramine (n = 8); the SSRIs comprised fluvoxamine (n = 14) and paroxetine (n = 19). In TCA-treated patients, the authors observed the same type of abnormalities in conduction and orthostatic hypotension as had been observed earlier. The authors also observed cases of first-degree atrioventricular block, prolonged QTc interval, and orthostatic hypotension in SSRI-treated patients. Thus SSRIs also appear to affect the cardiovascular system, which might pose a problem for patients with preexisting conduction disease. The authors observed a strong correlation between the decrease in systolic pressure and antidepressant serum concentration (except for clomipramine and paroxetine), suggesting that antidepressant serum level is a better correlate than dose.