The prevalence of infection with the genital herpes simplex virus type 2 (HSV-2) has been difficult to ascertain, primarily because of the large percentage of subclinical cases and the limitations in specificity of serologic assays for antibody to HSV-2. To obtain an improved estimate of the distribution of HSV-2 infection in the United States, we used an HSV type-specific antibody assay to test serum samples from 4201 participants in the second National Health and Nutrition Examination Survey. The results in our sample indicate that in the period from 1976 to 1980, 16.4 percent of the U.S. population 15 to 74 years of age (approximately 25 million persons) was infected with HSV-2 (95 percent confidence interval, 14.2 to 18.6 percent). Age and race were the demographic factors associated most strongly with the presence of HSV-2 antibody. The prevalence of the antibody increased from less than 1 percent in the group under 15 years old to 20.2 percent in the group 30 to 44 years old; it increased only slightly thereafter. In the oldest group, 60 to 74 years of age, the prevalence was 19.7 percent in whites and 64.7 percent in blacks. Among blacks of all age groups, but not whites, higher rates were observed in women than in men. The associations were weaker with respect to marital status, income, education, urban residence, and region of the country. After control for age, sex, and race, only the association with marital status remained significant; the rate was increased in persons previously married--i.e., divorced, separated, or widowed. We conclude that the prevalence of HSV-2 infection in the United States is higher than has previously been recognized and that many infections with this sexually transmitted virus may be subclinical.