Primary West Nile virus (WNV) infections can be diagnosed using a number of tests that detect infectious particles, nucleic acid, and specific IgM and/or IgG antibodies. However, serological identification of the infecting agent in secondary or subsequent flavivirus infections is problematic due to the extensive cross-reactivity of flavivirus antibodies. This is particularly difficult in the tropical Americas where multiple flaviviruses cocirculate. A study of sequential flavivirus infection in horses was undertaken using three medically important flaviviruses and five widely utilized diagnostic assays to determine if WNV infection in horses that had a previous St. Louis encephalitis virus (SLEV) or dengue virus type 2 (DENV-2) infection could be diagnosed. Following the primary inoculation, 25% (3/12) and 75% (3/4) of the horses mounted antibody responses against SLEV and DENV-2, respectively. Eighty-eight percent of horses subsequently inoculated with WNV had a WNV-specific antibody response that could be detected with one of these assays. The plaque reduction neutralization test (PRNT) was sensitive in detection but lacked specificity, especially following repeated flavivirus exposure. The WNV-specific IgM enzyme-linked immunosorbent assay (IgM ELISA) was able to detect an IgM antibody response and was not cross-reactive in a primary SLEV or DENV response. The WNV-specific blocking ELISA was specific, showing positives only following a WNV injection. Of great importance, we demonstrated that timing of sample collection and the need for multiple samples are important, as the infecting etiology could be misdiagnosed if only a single sample is tested.