The 86-item Midi-Mult was compared with the full MMPI in two studies. In Study 1, the Midi-Mult was extracted from the MMPI answer sheets for 301 patients. Correlations between comparable scales ranged from .65 to .89. Most scales showed significant differences in T-score elevations between the two forms, although the magnitudes were small. High-point codes derived from each form yielded good agreement in less than one quarter of the cases, while agreement improved with increasing pathology. In Study 2, the Midi-Mult was both extracted from the MMPI and administered separately. Correlations and high-point code agreement for 15 psychiatric patients were generally lower with the separately administered form. It was concluded that the Midi-Mult could not be used as an alternative to the MMPI for most clinical applications.