The criterion validity of the Toronto Alexithymia Scale (TAS) was assessed by administering the scale to 46 patients referred to a behavioral medicine outpatient clinic. Clinical ratings derived from observed interviews served as the criterion. TAS scores were significantly higher for the group of patients identified by two out of three raters as "alexithymic" than for the group identified as "nonalexithymic." On the basis of these findings, preliminary TAS cutoff scores were suggested. The results from this study and from previous investigations assessing the reliability and construct validity of the TAS indicate that it is currently the psychometrically best-validated measure of alexithymia. Further refinement and cross-validation with other clinical samples are recommended.