A principal components analysis of the 550 MMPI items on 1,576 patients referred for coronary angiography was performed to define the MMPI's psychological content dimensions and evaluate the comprehensiveness of its items. After failing to replicate the 21-factor solution of J. H. Johnson et al (1984), the authors interpreted 9 orthogonally rotated components. Agreement was found between a number of studies on several factors, including neuroticism, somatic complaints, cynicism, and religious orthodoxy. However, only 1 of the 5 personality dimensions identified by W. T. Norman (1963) was adequately represented. The 9-factor scales identified in this analysis may be useful as an alternative way of scoring the test, especially for longitudinal studies with archival MMPI data sets. It is suggested that the MMPI is capable of making subtle diagnostic distinctions when applied to a psychiatric population but that these distinctions become less meaningful when applied to a normal sample. Thus, it would be advisable to supplement the MMPI with instruments that measure a broader range of normal personality characteristics.