BACKGROUND: Seroprevalence of herpes simplex virus type 1 (HSV-1) and HSV-2 was determined in 1993 and 1998 in a randomly selected study group of 1024 and 654 attendees, respectively, at the sexually transmitted disease (STD) clinic of the University Hospital Rotterdam-Dijkzigt, The Netherlands. Correlations of HSV-1 and HSV-2 seropositivity were investigated. The relationship between HSV-1 and HSV-2 antibodies was also studied. METHODS: Data were collected in a cross-sectional study from February 1993 until February 1994 and from January 1998 until December 1998. Glycoprotein G (gG) HSV type specific serum IgG was determined. RESULTS: Seroprevalence of HSV-1 was 68% versus 59% (1993 versus 1998, chi(2)-test P < 0.001), of HSV-2 it was 30% versus 22% (1993 versus 1998, chi(2)-test P < 0.001). Using logistic regression analyses, HSV-1 and HSV-2 seropositivity were significantly associated with age and ethnicity in both groups. In 1993, HSV-1 seropositivity also correlated with lower level of education and female gender, whereas in 1998 it correlated with 'number of sexual partners in the past 6 months' and 'present diagnosis of STD'. In both groups, HSV-2 seropositivity was also more prevalent in females and related to sexual lifestyle variables. In an exposure-disease model, HSV-1 seropositivity was not correlated with HSV-2 seropositivity (odds ratio 1993 = 1.1, 95% CI : 0.8--1.7; odds ratio in 1998 = 1.0, 95% CI : 0.5--1.8). CONCLUSIONS: Seroprevalence of HSV-1 and HSV-2 is falling among STD clinic attendees in Rotterdam. A changing pattern of risk factors for HSV-1 seropositivity indicates increasing sexual transmission of HSV-1. Seropositivity for HSV-2 correlated with known risk factors. A previous HSV-1 infection does not reduce susceptibility to subsequent genital HSV-2 infections.