We determined the extent of the serologic cross-reactivity in the indirect fluorescent antibody (IFA) test for Babesia gibsoni, and the optimal cut-off titer for seropositivity in the test. The lowest titer to B gibsoni detected in a dog with naturally acquired clinical babesiosis was 1,280, and 7 of 12 dogs had titer between 10,240 and 20,480. Two experimentally infected normosplenic dogs developed high titer (40,960 to 81,920) to B gibsoni, and the same sera reacted in IFA tests for B canis (titer < or = 640), Toxoplasma gondii (titer < or = 2,560), and Neospora caninum (titer < or = 10,240). Dogs that were experimentally infected with B canis and T gondii had titer < or = 160 to B gibsoni. Dogs that were experimentally infected with N caninum had titer (80 to 10,240) to N caninum, but failed to have serologic reactivity to B gibsoni. Serologic titer of healthy dogs from Australia, a country where B gibsoni is not known to exist, was < or = 160 to B gibsoni. On the basis of these findings, a cut-off titer of 320 was considered to be appropriate for serodiagnosis of B gibsoni in dogs with clinical signs of babesiosis. A more conservative titer of 1,280 was established as the cut-off titer for seroepidemiologic studies based on relative costs and benefits of false-positive results and failure to isolate B gibsoni parasites after splenectomy and immunosuppression from a clinically normal dog with B gibsoni titer of 640.(ABSTRACT TRUNCATED AT 250 WORDS)