This article shows that college performance and selectivity have signi
ficant effects on earnings. It suggests that work that does not includ
e college performance overstates the effect of college selectivity for
Whites and understates it for Blacks. While the size of the effect of
college selectivity for Blacks is larger than for Whites, the large B
lack earnings gain is offset for students whose own Scholastic Aptitud
e Test scores are significantly below the median of the college they a
ttend. This results from the lower probability of graduation for ''mis
matched'' Blacks and the subsequently lower earnings for those who fai
l to complete college.