We reexamined the Japanese version of the 20-item Toronto Alexithymia Scale, a self-report sc ale for measuring alexithymic characteristics, by comparing the scores on three factors and the total scores with variables of the Rorschach in a sample of 40 (originally 48) Japanese college students. Based on prior studies, our aims were to further validate the 20-item Toronto Alexithymia Scale by comparing its scores with those on a projective technique. We also investigated whether sociocultural factors, such as repression of hostility. are associated with scores on the 20-item Toronto Alexithymia Scale (especially Factor 3). None of the seven Rorschach Alexithymia Variables were significantly related to the factors and total scores of the 20-item Toronto Alexithymia Scale. However, scores for Factor 1 (difficulty identifying feelings) of the 20-item Toronto Alexithymia Scale correlated positively with scores on Sum C' (reserved responses to emotional stimuli) and Adj es (stimulus demand), suggesting that individuals who score high for Factor 1 experience gloomy, depressive feelings but in constricted ways. Scores for Factor 3 (externally oriented thinking) correlated positively with the D scores (stress tolerance) and negatively with m (situational stress) responses, suggesting that Factor 3 may reflect a psychological defense that enhances stress tolerance. There was a nonsignificant negative correlation between Factor 1 and Factor 3 scores, but, unlike our hypothesis. Factor 3 was neither related to AG (aggression) nor S (space responses reflecting oppositional tendency), indices of aggression or hostility in the Rorschach Comprehensive System. It may be that the 20-item Toronto Alexithymia Scale and the Rorschach measure quite different aspects of personality, but further research is necessary.