Using a national level sample survey on labour market in India, we analyze the role of education–occupation (mis-)match (EOM) in explaining within-group dispersion in returns to education. Applying a double sample selection bias correction and Mincerian quantile wage regression estimation, the analysis reveals interesting findings. First, on average, overeducated workers suffer a wage penalty of 7% and undereducated workers do not receive a wage reward as compared to their adequately educated counterparts. Second, the inclusion of match status reduces within-education group dispersion in returns. The finding highlights that ignoring EOM and thus, adopting a restrictive view of similarity across workers may lead to overestimation of the within-education group dispersion in returns. This study argues for focusing on EOM to increase both pecuniary and social benefits of education in terms of productivity gains and wages as well as to reduce wage dispersion. © 2020, Springer Nature B.V.