• In Atlanta, the birth prevalence of spina bifida declined from 10.1 per 10,000 live births in 1972-1973 to 5.8 per 10,000 live births in 1978-1979. This trend cannot be explained by use of prenatal diagnosis, because it was unavailable in Atlanta until 1976, and even in the late 1970s was not used widely. To determine if this decrease was associated with changes in the distribution of clinical characteristics among infants with spina bifida, we reviewed the medical records of a population-based group of 154 infants with spina bifida, born in Atlanta during the eight-year period from 1972 through 1979. Distribution by the highest level of the defect on the spine did not change during these years. The proportions of infants with isolated spina bifida (ie, no other major malformations) and with open spina bifida (ie, not covered by skin) did, however, decrease. When we examined these two characteristics simultaneously, we found declines in both the proportion of infants with open-isolated spina bifida and the birth prevalence of infants with open-isolated spina bifida. If this trend persists, it will have important implications for workers involved with prenatal α-fetoprotein screening and researchers investigating the etiology of spina bifida.