The aim of this study was to investigate the types and prevalence of disease-specific antinuclear antibodies (ANAs) and their relationship to rheumatic diseases in the general Japanese population. An immunofluorescence (IF) method was used for the first screening of ANA levels in serum samples obtained from 2181 residents of a small Japanese town. Individuals positive for IF-ANA were then further tested for disease-specific ANAs using eight enzyme immunoassays. Physical status and the presence of illness were determined by means of questionnaires and medical examinations. Based on the result of the IF-ANA assay, the rates of positive samples at 1:40 and 1:160 dilutions were 26.0 and 9.5%, respectively, with females have significantly higher positivity rates than males (P < 0.0001). Among 566 IF-ANA-positive individuals, 100 individuals were found to have 114 disease-specific ANAs. Anti-SSA/Ro, anti-centromere, and anti-U1RNP antibodies were detected in 58, 30, and 11 individuals, respectively, but anti-Sm, anti-Scl-70, and anti-Jo-1 antibodies were undetectable. Questionnaires and medical examinations revealed that among 60 disease-specific ANA-positive individuals that were available for testing, six had Sjögren's syndrome (SS), five were suspected of having SS, and five had rheumatoid arthritis. Surprisingly, 34 (57%) of the disease-specific ANA-positive individuals were clinically healthy. Anti-SSA/Ro, anti-centromere, and anti-U1RNP antibodies were quite frequent among clinically healthy Japanese subjects, although anti-Sm, anti-Scl-70, and anti-Jo-1 antibodies were not. Of the 60 individuals who tested positive for disease-specific ANAs, 30% (18/60) actually manifested systemic rheumatic diseases, while 50% showed no detectable signs or symptoms of rheumatic diseases.