The Ordinary Least Square coefficient of years of schooling in standard Mincer’s (1974) equation may not reflect the effect of education on wages if it is correlated with ability of individual that also correlated with wages. Furthermore, productivity and skills between gender also may contribute differently on wages. Hence, this paper aims to estimate the economic returns to schooling on wages for Malaysia by using data gathered from the field survey of 1263 respondents in 2015. The contribution of this paper is twofold. First, we construct an ability index of individual during schooling and workplace to tackle the problem of omitted ability variable bias in OLS estimation. Then we compare this outcome with instrumental variable method by using family background as the instrument for education, taking into account the endogeneity problem. Second, we also investigate if the overall effect of education on wages reponds differently between genders by comparing both OLS and IV estimations. Our results indicate that family background is proven to be the best instrument for education, taking into account the endogeneity problem. Then, the IV rate of return for overall sample and split sample are higher than OLS. In regard of genders, the OLS suggests higher rate of return of men while IV estimates higher rate of return of women. These results have particularly important policy implications, concerning the government wage policies to achieve fair income between genders and thus contribute to the further development of Malaysian labor market efficiency.