Addressing certain problems with the compositional structure of the self-report Toronto Alexithymia Scale, this paper reports the development of a revised version of the scale as well as results from a preliminary series of studies evaluating its reliability and validity. The Revised Toronto Alexithymia Scale (TAS-R) demonstrated good internal consistency and a stable and replicable two-factor structure that is congruent with the two major dimensions of the alexithymia construct. Evidence of convergent and discriminant validity for the TAS-R was provided by correlations with measures of other constructs. Criterion validity was demonstrated by the ability of TAS-R scores to discriminate between behavioral medicine outpatients designated alexithymic and those designated non-alexithymic on the basis of clinical interview ratings. Normative data are provided for university student, normal adult, and psychiatric outpatient populations.