The purpose of this study was to examine the factor structure of the Frost Multidimensional Perfectionism Scale (MPS-F; Frost, Marten, Laharte, & Rosenblate, 1990). Although perfectionism is thought to contribute to the development of psychopathology and the MPS-F is gaining popularity for use in assessing perfectionism in clinical samples, to date the factor structure has not been examined in a clinical sample. Three hundred and twenty-two individuals diagnosed with an anxiety disorder using the SCID for DSM-IV and 49 nonclinical controls completed the MPS-F as well as a measure of perfectionism (MPS-H) developed by Hewitt and Flett ( 1991 ). Analyses suggested that the MPS-F has similar psychometric properties in clinical samples to those in nonclinical samples, and factors very similar to those observed by Frost et al. (1990) could be extracted. A 3-factor solution appeared more appropriate for statistical reasons, and the 3 scales based on these factors distinguished among diagnostic groups in a manner similar to scales based on the 6-factor solution in past research. Results were discussed in terms of the potential utility of a 3-factor solution and in terms of the general construct of perfectionism and the distinction between nonpathological high performance standards and neurotic perfectionism.