This study examines whether one’s religious orientation leads her to discriminate against members of other religious groups; whether this discrimination is affected by the exposure to others with different religious orientation; and how this affect human capital formation. These questions are especially relevant in light of the surge in religious tension around the world. We use data on national matriculation exams in Israel’s Jewish state education. Each exam is examined by two independent randomly assigned examiners who do not know the identity of the student. The religiosity of students is revealed to examiners by a special notation used by religious Jews. We identify the religiosity of examiners by the type of schools their children attend, secular or religious. We compare grades given to religious versus secular students by religious versus secular examiners. We estimate heterogeneity of this bias by examiner characteristics and environment at home and at work. We find evidence of in-group religious biases of examiners. These biases are especially large and significant among male examiners. These in-group biases have long-run implications since they affect students’ grades and chances of receiving a matriculation certificate, leading to allocative inefficiency. In addition, we find that in some cases exposure to others with different religious beliefs reduce the degree of these in-group biases, especially among male examiners.