In this study, sera from 116 children without clinical evidence of cat scratch disease (CSD) and sera from 19 children with CSD were investigated by two different immunofluorescent assays (test A and test B) to evaluate the seroprevalence of Bartonella henselae antibodies in this population. Antibodies against Bartonella henselae were found in noninfected children only in low titers: 13% and 3.5% of serum samples were positive by test A and test B, respectively. IgG titers as determined by test A ranged from 1:64 to 1:256, whereas test B did not yield titers >1:64. Of the 19 children with CSD, 7 (37%) and 5 (26%) had low antibody titers (1:64-1:256) on admission by test A and test B, respectively. In such cases, IgM antibodies against Bartonella henselae and/or a titer increase in the following weeks are required to prove suspected CSD. The results indicate that the seroprevalence of Bartonella henselae antibodies in children without evidence of CSD is low, and that the results may differ depending on the tests being used. Low antibody titers, however, were found not only in healthy children but also among patients with CSD, which could be indicative of the beginning or of the end of an illness. Currently, only few data are available concerning the seroprevalence of antibodies to Bartonella spp. among children. Although children are often affected by CSD, the specificity of most tests has been evaluated by investigating blood donors, who are usually adults. Several different serological tests are used for the diagnosis of CSD, including the immunofluorescence assay, the enzyme-linked immunosorbent assay, and Western blot (WB) analysis. The sensitivities of different IFAs range from 14 to 100%, depending on the antigen used, the cut-off chosen, and the test procedures. The current diagnostic value of different serological tests for diagnosis of CSD is reviewed.