A 2-year, three-panel (Tl-T3) longitudinal study of 233 entering Ph.D. students examined the relationships between student potential for mentoring, (i.e., attitudes and objective abilities at entry (Tl), mentoring functions used by the faculty adviser (T2, T3), and student research productivity and commitment (T3). Student potential was found to predict the amount of psychosocial mentoring, career mentoring, and research collaboration provided by the adviser. Psychosocial mentoring and collaboration were not related to student productivity or commitment after controlling for the students' entering abilities and attitudes. Career mentoring at T2 was negatively related to the students' affective commitment to their program at "3. Implications for our understanding of mentoring and future research are discussed.A growing body of research argues that mentoring is related to positive outcomes for the protCgC and the organization such as greater commitment, better socialization, better performance, higher salaries, and promotions (see for example, Chao, Walz,