Depressive illness has been reported to interfere with effortful processing, which requires conscious attention. The aim of this study was to evaluate divided attention in depressed patients, as a function of the degree of difficulty of the task performed. Tasks designed to measure unimodal and bimodal reaction times were presented to 10 patients with major depression and 10 normal control subjects. Performance was evaluated both before treatment when the patients were depressed and after treatment when they had recovered. Unlike the unimodal trials, the bimodal reaction time tasks were designed to evaluate decision-making under conditions in which attention was divided between two perceptual channels. Reaction times were measured under two different conditions in order to assess the extent of the response delay induced by divided attention, modality shifting, and decision processing. During simple response tasks, the depressed patients displayed significantly greater lengthening of reaction times when their attention was divided between two perceptual channels. This cross-modal delay effect occurred both for stimuli of the same modality and when shifting between modalities. The cross-modal delay effect was evident only for the choice tests in both the depressed and the recovered patients, but only the recovered patients were as accurate as the control subjects. These results suggest that the need for decision processing in depressed patients results in a failure to allocate the mental resources required to complete interchannel shifting, when attention is divided between two perceptual channels. These data are consistent with the hypothesis that attentional regulation is impaired in major depression.