This is a prospective screening study addressed to assess the value of femur and humerus shortening in the prenatal detection of Down syndrome. Prior to amniocentesis, 1,543 consecutive pregnancies between 13 and 18 weeks were studied. Femur and humerus shortening were assessed with the use of 6 different ratios, and then correlated with the karyotype obtained in amniotic fluid. Sensitivities achieved for Down syndrome with femur ratios were lower than those using humerus (17-22 vs. 43%) for similar false-positive rates (7-8 vs. 6-8%). The most effective ratio was the observed-to-expected humerus length with 43% sensitivity for a 6% false-positive rate. Combining femur and humerus measurements did not substantially improve the prediction obtained using the best humeral ratio. In conclusion, femur and humerus shortening appears to be of limited value in the detection of Down syndrome.