A series of interrelated analyses were conducted on 2 samples of college students to examine the reliability and validity of the Tridimensional Personality Questionnaire (TPQ) and to develop and validate a short version of the scale. Factor analyses were conducted and tended to approximate Cloninger's proposed model. Novelty Seeking predicted a range of substance use and abuse measures, and substance use disorders. Harm Avoidance was unrelated to substance use but predicted alcohol problems and dependence phenomena. No support was found for the quadratic and specific interaction effects proposed by Cloninger (1987a) . The Short-TPQ, a 44-item scale derived through exploratory factor analysis, demonstrated levels of reliability and validity quite comparable to those of the full scale. These results provide moderate support for the reliability and validity of both the TPQ and the Short-TPQ.