Abstract
An analysis is made of alphafetoprotein (AFP) concentrations in 3630 amniotic fluids submitted for prospective prenatal diagnoses over a 7‐year period. There were 89 cases of anencephaly, 74 of open spina bifida and 3467 with normal singleton outcomes. The AFP data were expressed in both standard deviations above the mean and multiples of the normal median for individual weeks of gestation. False positive and false negative rates were comparable in the two systems at selected cut‐offs. It is concluded that either system may be used in setting action limits for the primary distinction of unaffected pregnancies from those in which an open neural tube defect is present.