Theoretical frameworks and experimental paradigms derived from cognitive psychology provide a valuable approach to the investigation of cognitive aspects of disorders of anxiety and depression. Review of the literature provides strong evidence of attentional bias towards threat stimuli in anxiety, though evidence of comparable attentional bias in depression is weaker. There is also strong evidence of memory bias towards mood-congruent materials in depression, though comparable evidence regarding anxiety-congruent memory is mixed. A variety of theoretical frameworks for explaining such phenomena have been advanced, using concepts such as networks, schema and mental models. There is a need now for theoretical frameworks that address themselves to the specificity of the empirical phenomena. Clinical and empirical implications of these findings and models are discussed, including developments in diathesis-stress models of psychopathology, the use of experimental-cognitive paradigms in clinical assessment, and the prospect of therapeutic applications. © 1990.