To verify and investigate the reported decline of neural tube defects (NTDs) in the last decade, data from three different sources were examined: the Birth Defects Monitoring Program (BDMP), composed of newborn discharge diagnoses; the population-base Metropolitan Atlanta Congenital Defects Program (MACDP); and the National Center for Health Statistics (NCHS), whose data include only live births. Although the three systems use different methods of ascertaining cases, all data bases showed a decreasing trend in NTD rates, with annual percent decreases of 3.1% for NCHS, 5.7% for BDMP, and 7.7% for MACDP. The decrease was noted in all variables examined: race, sex, and birth status for each defect. NTD rates appear to be declining more rapidly in females than in males. Further, the rate among stillborn infants decreased more than the rate among live-born infants. The exclusion of stillbirths in the NCHS data may contribute to its low rate of decline for female anencephalics. Data from England showed a larger average annual decrease (10.6%) than did data from the United States. Although reasons for the overall decline are not verifiable, a variety of explanations are suggested for this decline and for the difference in NTD rate decline noted between live births and stillbirths. The overall decline in NTD rates should be investigated with respect to etiology; recognition of the decline is useful for planning and evaluating health care programs.