A previous paper described the development of the twenty-item Toronto Alexithymia Scale (TAS-20) and reported preliminary evidence of reliability and factorial validity of the scale. This paper describes a study that further evaluated the construct validity of the TAS-20 by examining its relationship with measures of personality traits theoretically related or unrelated to the alexithymia construct, as well as its relationship with an observer-rated measure of alexithymia. Evidence of convergent and discriminant validity of the TAS-20 was demonstrated in samples of university students by a pattern of correlations with the scales of the NEO Personality Inventory and separate measures of psychological mindedness and need-for-cognition that was consistent with theoretical predictions. The concurrent validity of the scale was demonstrated by positive correlations with observer-ratings of alexithymia in a sample of behavioural medicine out-patients.