Standard regression equations relating the total fertility rate to contraceptive prevalence indicate that the fertility rate of 5.5 lifetime births per woman observed for Northeast Brazil in the 1986 Brazil Demographic and Health Survey is about 1.6 births per woman higher than would be expected on the basis of contraceptive prevalence at that time. An exploratory approach for evaluating the reasons for higher-than-expected levels of fertility attributes 0.6 of the apparent excess births in Northeast Brazil to the lag effects of recent increases in contraceptive use and 0.6 to the relatively small fertility-inhibiting effect of breastfeeding. Marriage patterns play a smaller role, but appear to be responsible for 0.3 births of the remaining difference between observed and expected fertility.