PIP: Data from the 1986-90 Demographic and Health Surveys of Burundi, Ghana, Kenya, Mali, Nigeria, Senegal, Togo, and Uganda were used to examine the impact of fertility, child mortality, and socioeconomic and demographic factors on female rural-urban migration of six months or more duration. Several principles appear to direct the migration of high fertility women. High parity women are free from the male demands for more children. Rural areas lack basic amenities such as schools, health services, and modern housing. Additional children may strain family resources and require additional income from other sources. Husband and kin may have already moved. All data are nationally representative, with the exception of Uganda with an 80% sample. Women living in rural areas two years prior to the survey were included in the sample. Fertility and mortality data pertain to children aged under five years in the period two to seven years before the survey year. Multinomial logit analysis was based on an analytical model developed by Goldstein and Goldstein. High fertility was found to deter female migration to either urban or rural areas. Women, who had one surviving child aged under five years, were significantly less likely to move to urban areas only in Nigeria and Senegal and to other villages in Burundi compared to women without a recent birth. Women with two or more surviving children were significantly less likely (by 43-75%) to move to urban areas in five out of seven countries. Moves to rural areas were less likely by 36-61% in six out of eight countries. The evidence does not suggest that the reason for moves is to advance the children's material or physical well-being. The number of births, particularly in Kenya where fertility is very high in rural areas, acts as a deterrent to migration. Child mortality only constrains moves to urban areas. Unmarried women, single women, better educated women, and adults in their 20s are more likely to move to urban areas.