The aim of this study was to predict the long-term protection induced after immunisation with inactivated, aluminium-free virosome hepatitis A vaccine. The study population consisted of adult volunteers enrolled in four different clinical trials. Lower 95% confidence interval limits and seroconversion rate were calculated by using a linear mixed model to estimate the persistence of serum antibodies over time. To assess the robustness of the mathematical model, several sensitivity analyses were performed with more conservative protective threshold (20 mIU/ml vs. 10 mIU/ml), higher yearly decline rate, and exclusion of volunteers who had increasing titres over time. Based on 190 volunteers with at least two valid assessments of titres from year 3 onward, the median duration of protection was 55.5 years, with a lower limit of the 95% CI of 48.7 years. Duration below 25.3 years was predicted for only 5% of the subjects. Women tended to have higher titres to start with, but their rate of decline was higher, resulting in similar duration of protection overall. The use of a more conservative threshold, higher yearly decline rate, and exclusion of volunteers with increasing titres over time did not affect these results. According to this model, 95% of the volunteers should have anti-HAV titres above the minimum protective threshold for 20 years or more following immunisation with two doses of this aluminium-free vaccine.