This paper is an empirical attempt to quantify caste-based discrimination in the labor market using household data taken from rural North India. In the regression analysis, transaction costs associated with entry into the labor market and reservation wages are estimated simultaneously along with market wages. The estimation results provide evidence of the existence of transaction costs in the labor market and discrimination against backward classes with regard to access to regular employment. In line with previous studies, the results suggest that the achievements of India's reservation policy so far have at best been limited. In addition, a comparison between the estimates from the model employed in this paper and conventional (reduced-form) approaches shows that discrimination in labor market entry is likely to be underestimated in the conventional reduced-form approaches.