The present study investigated the unique contribution of learning approaches to academic performance, also taking gender differences into account. The participant sample consisted of 476 school pupils (53% girls and 47% boys) from two upper secondary schools in Sweden who completed two self-reported measures related to personality and learning approaches and one cognitive ability test. A series of hierarchical regressions were performed with participants' school subject-specific grades as the criterion variable and learning approaches as the predictor variable, after controlling for all variance related to IQ and personality. The results showed that learning approaches accounted for 6% and 16% of the variance in academic performance for girls and boys, respectively. The results are discussed in terms of possible explanations for and implications of the gender differences found. (Contains 4 tables.)