Social cognitive career theory (SCCT, Lent, Brown, & Hackett, 1994) represents a comprehensive conceptual framework that helps understand career choice and academic performance. These authors propose, following the triadic model of causation, a series of theoretical relationships and causal pathways (Bandura, 1987). They emphasize the role of self-efficacy beliefs, outcome expectations, and goals. Several recent empirical studies indicate that personality traits are (a) associated with variables such as motivation, intelligence and self-efficacy (Ackerman & Heggestad, 1997) and, perhaps more important, (b) are important predictors of academic performance (Chamorro-Premuzic & Furnham, 2003a). This has prompted educational researchers to explore the relationship between personality traits and SCCT. The purpose of this study is to extend the SCCT's mathematic performance model by clarifying the explanatory contribution of personality traits. For this purpose, the logical-mathematical self-efficacy scale, math outcome expectations, math performance goals and mathematics ability test and the Big Five Questionnaire–Children were administered to a sample of 543 adolescents (44% female and 56% male) who were enrolled in the eighth or ninth year of the final cycle of General Basic Education (EGB), aged between 12 and 16 years (M = 13, 96, DS = .84) and pursuing their studies in state schools (39.9%) or private (60.1%) of the city of Cordoba, Argentina. In the first study, multivariate analyses of variance (MANOVA) indicated significant effects of gender and educational placement. The interaction between these factors was also significant. The effect size, however, ranged from small to moderate; we decided, therefore, to test the model fit for the full sample. A hierarchical regression analysis revealed that personality traits, math skills, self-efficacy beliefs/ outcome expectations and performance goals explained 9%, 13%, 10% and 5% of the math academic performance variance, respectively. All the predictors combined explained 37% of the variance of math school performance. In the third study a path analysis evaluated the relationship among variables in the hypothesized model. The analysis also indicated an adequate fitness from the data to the hypothesized model (GFI = .99; CFI = .99; RMSEA = .04). Separate covariance matrices for gender and educational placement were used to conduct a multiple group analysis, and an adequate fitness was observed too. These results suggest that academic performance is better in those students with greater math skills, greater beliefs about their capabilities and more demanding performance targets. In addition, students who exhibit higher self-efficacy beliefs have higher math skills and greater openness to try new experiences. Also, students who intend to more demanding performance targets are those with higher self-efficacy beliefs, expectations of positive results and a more responsible behavior. We could not confirm that more emotionally stable students present higher self-efficacy beliefs and expectations of positive results, nor that those with higher mathematical ability show greater positive outcome.