This paper presents a reexamination of James Coleman et al.'s study of white enrollment losses from desegregating urban school districts over the period 1968-1973. New equations are estimated using a different measure of desegregation, additional explanatory variables, and modified samples. The earlier conclusion that desegregation has been a significant stimulant of white enrollment losses in the largest central city districts is supported, although this overall effect comes almost entirely from districts in which black-white contact exceeds a threshold level. In addition, the age structure of the district's white population and the geographical coverage of the district are significant in explaining white losses for some samples. © 1979.