The secondary antibody response to 5.0 microgram flagellin was studied by haemagglutination in 132 healthy or convalescent subjects given a primary challenge with 5.0 microgram flagellin from 1 to 44 months previously. The peak titre, expressed as total antibody, occurred at 2 weeks and was mainly immunoglobulin (Ig)G. The magnitude of the titre of total antibody was influenced predominantly by that of total antibody in the primary response (P less than 0.001), the interval between primary and secondary responses (P less than 0.005) and the subjects' age (P less than 0.05) and sex (P less than 0.08). Together these accounted for 23% of the variability observed in the secondary response, with total antibody titre in the primary response accounting for 11% of the variability. The titre of IgG antibody was likewise influenced by these four variables, but the influence of age or sex on IgG antibody was not statistically significant. In human vaccination programmes, choice of the appropriate interval between primary and booster inoculations could increase prophylactic effectiveness and, if two inoculations were to prove as effective as three, there would be reduced work and increased public acceptance. Moreover, the demonstrable capacity for responsiveness of aged and debilitated persons should encourage the wider use of appropriate prophylactic immunization in these groups.