This review examines studies that have addressed the presence of selective attentional biases in depression (clinical and subclinical samples) in several experimental cognitive tasks. Current data using eye-tracking techniques indicate that depression is characterized by the presence of sustained attentional processing towards negative information and absence of biases towards positive information. Available empirical evidence about the causal role of these biases on the onset and maintenance of depression suggests that these biases, in association with a ruminative style and ineffective emotional regulation strategies, could be stable vulnerability markers of depression. Clinical implications of these findings and future research in this field are discussed.