OBJECTIVES: Establishing and comparing race, ethnic, and gender-specific birth weight percentiles for gestational age is requisite for investigating the determinants of variations in fetal growth. In this study, we calculate percentiles of birth weight for gestational age for the total 1994-1996 U.S. population and contrast these percentiles by racial/ethnic and gender groups. METHODS: Single live births to U.S. resident mothers were selected from the 1994-1996 U.S. Natality Files. After exclusions, 5,973,440 non-Hispanic Whites, 1,393,908 non-Hispanic African Americans, 1,683,333 Hispanics, 80,187 Native Americans, and 510,021 other racial/ethnic groups were used to calculate distribution percentiles of birth weight for each gestational age for which there were at least 50 cases to calculate the 50th percentile and 100 cases to calculate the 10th percentile. RESULTS: Fetal growth patterns among the four U.S. racial/ethnic groups varied markedly and, across the gestational age range, there was considerable oscillation in the relative ranking of any one group's birth weight percentile value in comparison to the others. Males had relatively higher birth weight percentile values than females. The proportion of infants with a birth weight value less than 1994-1996 U.S. population's 10th percentile value of birth weight for their corresponding gestational age was 7.87 for non-Hispanic Whites, 15.43 for non-Hispanic African Americans, 9.30 for Hispanics, and 8.81 for Native Americans. CONCLUSIONS: While the factors underlying trends and population subgroup differences in fetal growth are unclear, nutrition, smoking habits, health status, and maternal morbidity are possible precursors for part of the variations in patterns of fetal growth. As prenatal care has been touted as a means to reduce the risk of fetal growth restriction at term, assuring the availability and accessibility of comprehensive prenatal care services is viewed as an essential corollary in the effort to improve fetal growth patterns in the United States.