Comparisons of the seroepidemiology of genital herpes simplex virus (HSV) infection within and between countries are hampered by variations in tests, methods and populations sampled. Differences in seroprevalence may partly reflect variability in diagnostic efforts and healthcare awareness, expectations and utilization. To allow comparison between surveys and to improve their performance, seroepidemiological studies should use validated HSV type-specific tests, report age-specific or age-adjusted prevalence and define the period of time over which samples were collected. Despite the difficulty of comparing studies, the prevalence of HSV-2 infection varies between developed countries. Among healthy adult populations, HSV-2 seroprevalence is higher in the USA than in Europe. Furthermore, HSV-2 seroprevalence varies widely among European countries. For example, in 1989 HSV-2 seroprevalence among pregnant women was reported to be 33% in Sweden compared with 8.3% in Germany. In some, but not all, countries, HSV-2 seroprevalence appears to be increasing. In the USA, the National Health and Nutrition Examination Surveys found that HSV-2 seroprevalence increased by almost one third from 16.4% to 21.8% from 1976 to 1994 in people over 12 years old. The incidence of HSV infection is a measure of primary infection. HSV incidence is difficult to quantify, partly due to unrecognized or asymptomatic infections. However, estimates of incidence in North American and European populations range from 5 to 24 per 100 people per year. Prevention programmes should recognize that HSV-2 seroprevalence increases rapidly in early adult life. The proportion of genital herpes infections caused by HSV-1 is increasing in the developed world, possibly due to changes in oral-genital sexual behaviour and lower rates of HSV-1 acquisition in childhood.