Prenatal diagnosis is now available to fragile X (fra[X]) syndrome families and has proven reliable when testing male fetuses. It has been reported that approximately one-third of heterozygous fra(X) females demonstrate mental impairment. Based on this, families usually continue pregnancies involving a female fetus. The purpose of this study is to investigate whether a 35% risk for mental impairment is appropriate when counseling heterozygous women carrying fra(X)-positive female fetuses. Forty-three cytogenetically positive (greater than or equal to 2%) daughters of known fra(X) carrier women were ascertained postnatally in an unbiased fashion. Their mother's carrier status was determined on the basis of at least one son with Martin-Bell syndrome. In addition to peripheral blood cytogenetic studies, all daughters had cognitive testing to determine full-scale IQ. In this study, 55.8% (24/43) were mentally impaired (IQ less than 85) compared with the expected 35%. Of these, 42% were mentally retarded (IQ less than 70). Although we do not know the correlation between percent fragility by peripheral blood compared with the percent fragility by amniocentesis or CVS, we assume that they are relatively comparable such that a female who is positive with greater than or equal to 2% fragility would probably be positive by all methods. This report suggests that the penetrance of mental impairment in females with a percent fragility of greater than or equal to 2% may be as high as 55%. Further studies are necessary to clarify this issue so that accurate information can be given in genetic counseling.