Current algorithms to determine eligibility for prenatal cytogenetic diagnostic services depend critically on the accuracy and precision of the underlying rates of cytogenetic abnormality used in the calculations. We examine the maternal age-specific rates of Down syndrome livebirths in eight studies of European-origin populations, pooled rates from which are widely used for baseline calculations in biochemical screening. These studies vary significantly in such factors as methods of ascertainment of cases, likelihood of complete ascertainment, and methods of correction, if any, for underascertainment. Restriction of analysis to those two studies among the eight whose methods suggest the greatest likelihood of complete ascertainment for Down syndrome generates rates significantly higher than those in widespread use. Confidence intervals about previously reported and currently derived rates indicate that even with large-scale data, there is considerable residual uncertainty in derived rates.