A principal factor analysis, conducted on a mixed psychiatric outpatient sample (N = 470), identified both common and specific dimensions underlying anxiety and depression. Although an initial single-factor extraction accounted for a significant proportion of variance in cognitive and symptom measures of anxiety and depression, a two-factor solution, in which anxiety and depression formed separate dimensions, proved to be the better solution. MANOVAS performed on pure depressed, pure anxious, and mixed anxious/depressed subgroups provided evidence of a specific cognitive profile for anxiety and depression. The mixed subsample evidenced greater severity, a mixed cognitive and symptom profile, and character traits that may indicate increased vulnerability to psychological disturbance. Results are discussed in terms of Beck's (1976) cognitive content-specificity hypothesis and the positive-negative affect model (Watson & Tellegen, 1985).