This study was designed to examine selective processing of emotional information in depression. It focuses on possible attentional biases in depression, and whether such biases constitute a cognitive vulnerability factor to suffer from the disorder or, on the contrary, they reflect a feature associated exclusively with the clinical level of depression. 81 participants were included in the study: 15 with a diagnosis of Major Depression; 17 were diagnosed as Dysthymia; 11 participants scored over 18 in the Beck Depression Inventory (Beck, Rush, Shaw, & Emery, 1979); 15 participants, in whom a sad mood state was induced by an experimental mood induction (Velten technique + music, or biographical recall + music); and 23 participants as a normal-control group. All participants were presented with the emotional Stroop task. The data indicated that attentional bias was only present in the group of patients with Major Depression, so it does not seem to be a cognitive vulnerability factor for this disorder.