The authors determined the seroprevalence of hepatitis B virus markers among Danish dentists in 1985 in an effort to assess occupational risk factors. A case-control study approach was applied using pedodontists as zero-exposure controls. The study further allowed serologic and epidemiologic follow-up of dentists who participated in a similar study in 1976. The authors sought to determine whether exposure to hepatitis B virus in this population sometimes leads to immunization without infection (antibody to hepatitis B surface antigen (anti-HBs) positivity only). A total of 922 dentists (22% of all Danish dentists) gave blood; of these, 254 (28%) participated in the 1976 study. Serum samples were tested using commercial solid phase radioimmunoassays. The prevalences of anti-HBs and antibody to hepatitis B core antigen (anti-HBc) were 7.2% (95% confidence interval (Cl) = 5.7-9.0) and 5.2% (95% CI = 3.8-6.6), respectively. Twenty-one dentists (31%) only had anti-HBs (without prior vaccination). Age-standardization of the two data sets (1976 and 1985) indicated no substantial change in seroprevalence. Private practitioners, especially those working in Copenhagen, had an increased odds ratio (OR) of being seropositive compared with that of pedodontists (OR = 3.0 (95% CI = 1.3-6.8) adjusted for age and sex and OR = 2.4 (95% CI = 1.1-5.2) adjusted for age and site of practice). The proportion attributable to occupational exposure was approximately 50%. In a logistic regression model, site of practice and type of specialty were associated with hepatitis B virus seropositivity (p = 0.01 and 0.05, respectively), whereas age, sex, hours of oral surgery per week, and use of gloves were not. Four persons had become seropositive in the follow-up study, for a seroconversion rate of 0.2% per year. Seven out of 18 dentists who were anti-HBs positive in 1976 were found to be anti-HBs negative in 1985. Most of these dentists had values below 10.0 units by S/N ratio (ratio of counts per minute of sample compared with the mean number of counts per minute of negative controls). The findings do not justify institution of a large-scale vaccination program for dentists, but there are certain subgroups of dentists which constitute candidates for vaccination.