Several studies have evaluated the value of cord blood IgE concentrations for predicting the risk of allergic disease in children. In all of these studies it is necessary to exclude cord blood samples in which the IgE may be falsely elevated due to admixing of maternal and fetal blood during parturition. The most common method for detecting mixing of fetal and maternal blood is measurement of cord blood IgA concentrations. We have examined the theoretical basis of IgA measurements for detecting maternal blood contamination and reexamined our own data to evaluate IgA measurements. Our data suggest that the average IgA concentration of 28.2 micrograms/ml used in previous studies may not adequately exclude contaminated cord blood sample. Failure to reject contaminated cord blood samples would reduce the positive predictive value of cord blood IgE measurements.