To estimate the risk of hepatitis B virus (HBV) infection among student nurses, we measured the prevalence of HBV infection in 218 student nurses (192 female and 26 male nurses) and the annual attack rate in a subgroup of 117 subjects. In both studies sera were tested with an enzyme immunoassay for the presence of HBsAg, anti-HBc and anti-HBs. In the prevalence study 24 student nurses (11.1%) had a marker of prior HBV infection; the presence of both anti-HBc and anti-HBs was the most prevalent serologic pattern. 2 student nurses were HBsAg positive (0.9%). Prevalence of infection was strongly related to increasing age, while the occupation of the head of the family showed no effect. In the incidence study 7 subjects seroconverted in a year (6.2%): 6 acquired only anti-HBc and 1 only anti-HBs. These data suggested the hypothesis that the presence of anti-HBc alone (a pattern usually associated with past infections in which anti-HBs is no more detectable) may represent false positive results. The implications of such hypothesis are discussed with particular reference to prevaccination screening policy.