High prevalences of anencephaly and neural tube defects (NTDs) have recently been recorded for several Texas counties bordering Mexico. In addition, a few investigators have reported Hispanics to be at elevated risk for NTDs (anencephaly and spina bifida). Factors contributing to this risk have not been established. The authors conducted a study of NTDs in Harris County, Texas, to determine the prevalence of each defect. Prevalence was established by identifying cases among resident live births and fetal deaths (stillbirths at > or = 20 weeks) occurring from April 1, 1989, through December 31, 1991. Using multiple case ascertainment methods, 59 cases of anencephaly and 32 cases of spina bifida were detected, resulting in prevalences of 3.8 (95% confidence interval 2.9-4.9) and 2.0 (95% confidence interval 1.4-2.8) per 10,000 live births, respectively. The ratio of anencephaly prevalence to spina bifida prevalence was 2:1 in 1989, 1:1 in 1990, and 3:1 in 1991, with a significant difference in 1991. The female:male prevalence ratio was 1.0 for spina bifida and 2.2 for anencephaly, and was higher still for anencephaly among non-Hispanics (prevalence ratio = 5.6). For each defect, Hispanics experienced a prevalence approximately three times that of non-Hispanics. This ethnic difference was greater for males with anencephaly and for females with spina bifida. For anencephaly, the Hispanic:white/Anglo prevalence ratio (4.2) and the African-American:white/Anglo prevalence ratio (1.9) were greatly elevated and the Hispanic:African-American prevalence ratio (2.2) was similar, relative to comparable studies from the past two decades. The prevalence of anencephaly recorded for public hospitals (7.0 per 10,000) was three times greater than that for private hospitals (2.4 per 10,000). Spina bifida figures were similar for public (prevalence = 2.2 per 10,000) and private (prevalence = 2.0 per 10,000) hospitals. A significantly higher prevalence of both defects was documented among Hispanics in Harris County. The higher anencephaly rates among Hispanics, African-Americans, and those using public hospitals in an era of NTD screening, prenatal diagnosis, and elective pregnancy termination suggest that socioeconomic and perhaps cultural/religious factors might influence the recorded birth prevalence of this defect in particular groups.