Humoral antibody response to three types of rabies vaccines were assayed by the neutralization (NT), the mixed hemadsorption (MH), and the indirect immunofluorescence (IF) tests. The NT and MH tests were used to detect antibodies combining with antigens at the surface of virions and infected cells, whereas the indirect IF test measured antibodies mainly to the rabies nucleocapsid antigen. After immunization with a human diploid cell vaccine, antibodies were detected by both the NT and the MH test in the 14th- and 30th-day serum samples from each of eight vaccinated persons. There was a good correlation between titers obtained with the two tests in this group of vaccinees. Antibodies elicited by duck embryo and nervous tissue vaccines occurred less frequently and in lower titers. In these groups of vaccinees, 5 of 14 and 5 of 10, respectively, had antibodies detectable by the NT test in the 14th- and 30th-day sera but were negative by the MH test. It is suggested that this was due to the high levels of immunoglobulin M antibodies, which are known to be elicited by daily injections of vaccine. Since antibodies of the immunoglobulin M class are considered to be less important for protection against rabies, the MH test is recommended for immunity determinations. Compared with the NT test, this test also offers the advantage of being technically more convenient because of its capacity for testing numerous sera in a single run. Antibody titers obtained by the indirect IF test in the human diploid cell vaccine group were relatively low. Titers in the duck embryo and nervous tissue vaccine groups were higher but did not correlate with the results of the NT test.