A prospective study of the diagnostic accuracy of ultrasound in the prediction of fetal microcephaly was performed on a study population of 24 patients. An occipitofrontal diameter larger than the predicted mean −2 standard deviations (SD), a head perimeter larger than the predicted mean −2 SD, and a head perimeter/abdominal perimeter larger than the predicted mean −1 SD were found to exclude fetal microcephaly. An occipitofrontal diameter smaller than the predicted mean −4 SD, a head perimeter smaller than the predicted mean −5 SD, a head perimeter/abdominal perimeter smaller than the predicted mean −3 SD, and a femur length/head perimeter larger than the predicted mean +3 SD were found to cause no errors in the diagnosis of microcephaly. If neither of these two groups of tests is satisfied, fetal microcephaly cannot be reliably diagnosed or excluded on the basis of a single ultrasound examination.