Recently, the alexithymia construct has emerged as a heuristically useful personality construct in the explanation of the pathogenesis of a variety of physical illnesses. Research that has evaluated the alexithymia construct, however, has been conducted with little attention to the psychometric properties of the various scales constructed to measure it. This study investigated the underlying dimensions and factor structure stability of the MMPI Alexithymia scale (MMPI-A). The MMPI-A was administered to 552 undergraduates, and the data then were split randomly into two samples. Factor analysis of both samples produced a three-factor solution with little theoretical relevancy to the alexithymia construct. Comparisons of the factor structures from the two samples revealed only moderate congruence, which indicates an unstable factor structure.