The cardiovascular effects and toxicity of tricyclic antidepressants (TCAs) have been well documented in medical literature. The most common manifestation of such effects is slowing of intraventricular conduction, manifested by prolonged PR, QRS and QT intervals on the standard electrocardiogram (ECG) and postural hypotension. In contrast to TCAs, selective serotonin reuptake inhibitors (SSRIs), including fluoxetine and citalopram, are considered to cause less effect on cardiac impulse conduction. In addition, these compounds induced significantly less anticholinergic, antihistaminergic and cardiotoxic side-effects than TCAs. However, there is an increasing number of case reports on dysrhythmias, like atrial fibrillation or bradycardia and syncope associated with fluoxetine and another SSRI treatment and overdose. Although such reports have not been common, they do raise concerns. In cardiac tissues isolated from canine, rabbit, rat and guinea pig hearts we have found that fluoxetine and citalopram inhibited cardiac Na+ and Ca2+ channels. These direct cardiac electrophysiological effects were similar to those of observed for tricyclic antidepressants clomipramine and imipramine. The inhibition of cardiac Ca2+ and Na+ channels by fluoxetine may explain most cardiac side-effects observed occasionally with the drug and mild but significant bradycardia reported during chronic treatment. Our results suggest that fluoxetine and citalopram may have antiarrhythmic (class I + IV type), as well as proarrhythmic properties (due to impairment of atrioventricular or intraventricular conduction and shortening of repolarization). Taking all these into consideration, in depressed patients having also severe cardiac disorders, ECG control may be suggested during fluoxetine and probable another SSRI therapy. The primary goal of this review is to compare these direct cardiac effects of fluoxetine and citalopram to those of previously reported for TCAs. This paper also summarizes the recently observed effects of fluoxetine apparently not related to the blockage of 5-HT transporter based on literature.