The empirical data came from the National Crime Panel between 1973 and mid-1977, which provides in addition to standard social and demographic measures of households, information on whether the household moved after an interview, whether a member had been a crime victim, the race of the households in units contiguous to the respondent, etc. In the sample of some 25,000 households in metropolitan areas, no 'flight' by whites -- or blacks -- was found. Simple residential mobility was higher in integrated areas. However, once characteristics of the households, particularly family income, number of young children, and home ownership, were accounted for, the mobility difference between households in integrated and segregated areas vanished. Crime victimization of a household member, on the other hand, did increase household residential mobility among white homeowners and blacks, but not among white renters. The pattern of racial change of the sampling clusters of four adjacent housing units showed little evidence that integrated clusters tended to become more black over periods of up to 30 months. Instead of avoiding these areas, whites seemed to move in at rates consistent with their proportion in the housing market. These findings suggest that there is little impediment to racial residential integration at the household level. It appears that the continuing high level of racial segregation in housing is caused by certain mechanisms, such as discriminatory real estate practices, rather than racially motivated actions by white households. Tables, footnotes, and over 50 references are supplied. Further data analyses are appended. (Author abstract modified)