This study examined whether motivational and psychosocial factors could predict persistence and performance longitudinally. Questionnaire data from the HELT (Helse- og trivsel blant studenter ved Universitetet i Oslo) study, a survey of students health and well-being at the University of Oslo in 2003, was used to predict the academic careers of 646 students over the course of two and a half years (from 2003 to 2005). Logistic and multiple linear regression analyses were conducted to investigate the relative contribution of academic motivation, social support, attendance, degree of full-time studies, mental health and personality on persistence (or dropout) and average grade. Measures included the Academic Motivation Scale, the Emotional versus Social Loneliness Scale, the HSCL-25, the 5Pfa (a short version of a Norwegian translation of Big Five (Engvik, 1993), and study related variables. Outcome variable data was collected from the student record at the University of Oslo. The results revealed different patterns of predictors for the two different outcome variables, as well as gender differences in the predictors. Attendance predicted persistence uniquely for both genders, and mediated the effect of amotivation. Female students persistence was further predicted by extrinsic motivation, emotional loneliness and personality (openness, agreeableness and neuroticism). Female students with a higher degree of extrinsic motivation and neuroticism, and a lower degree of emotional loneliness, openness and agreeableness had a higher probability of persisting towards completing a degree.The variables that had unique effects on grades for women were intrinsic motivation, amotivation, friends at the university and extraversion. Female students with higher intrinsic motivation, a larger social network at the university and lower amotivation and extraversion in 2003, achieved better grades throughout their academic careers. Amotivation and extrinsic motivation were negatively related to grades for men, but these effects were mediated by conscientiousness, which was the only predictor that contributed uniquely to the prediction of grades for the male students.Limitations of the current study, implications of the results and directions for future research are discussed.