The historic motivation for federal intervention in local school desegregation has been twofold: to improve educational conditions and human capital acquisition for black students, and to improve intergroup relations. Empirical evidence on the extent to which desegregation affected such relations is scant, however. We examine interracial birth rates as a revealed preference measure of the quality of intergroup relations and use variation in exposure to desegregated schools across cohorts arising from the idiosyncratic timing of court-ordered desegregation plans to identify the effects. Some point estimates suggest small positive effects on the share of births to black fathers where the mother is white, but the estimates are statistically insignificant. The estimates are sufficiently precise to rule out socially significant effects of school desegregation on interracial births; at the upper bound of the 95 percent confidence interval, school desegregation would explain about 15 percent of the substantial increase in mixed-race births between 1968 and 1995.