Formalin-inactivated, alum-adsorbed, hepatitis A vaccine was evaluated in 100 healthy adults who were stratified at enrollment into two age groups: 18-39 years: n = 50; 40-65 years: n = 50. All individuals received vaccine at 25 U of viral antigen. After stratification, both groups were randomized to receive either vaccination at 0 and 24 weeks or vaccination at 0.2 and 24 weeks. Subjects were bled for serology at 0, 2, 4, 24, 28 weeks and 1 year. The seroconversion rate and geometric mean titer (GMT = mIU ml-1) after one dose of vaccine was lower for older subjects [second week: < 40 years: 15/25 (60%) (GMT: 12.9). > 40 years: 5/22 (23%) (GMT: 6.1): fourth week: < 40 years: 20/22 (91%) (GMT: 29.0), > 40 years: 16/23 (70%) (GMT: 14.3)]. After a second dose at 2 weeks the seroresponse improved so that there were no longer differences between age groups [24 weeks: < 40: 21/22 (95%) (GMT: 123.9), > 40: 22/23 (96%) (GMT: 106.1)]. A third dose at 24 weeks resulted in a 20-40-fold increase in GMT in both age groups. As a separate evaluation height, weight, skin fold thickness, and body mass index (BMI) were assessed by logistic regression for their ability to predict serologic response. Serologic response was significantly associated with lower weight (P = 0.032) and BMI (P = 0.024) but not with height or skin fold thickness. Hepatitis A vaccine was well tolerated, with no serious adverse experiences. Adults older than 40 years appear to respond less well than younger adults to a single dose of 25 U of hepatitis A vaccine but equally well after two doses of vaccine. The slower antibody response to hepatitis A vaccine in overweight individuals was not attributable to skin adipose tissue.