Three diagnostic assays for detecting West Nile virus (WNV) in avian oral swabs were evaluated in California in 2004 and 2005: two commercial antigen-capture assays, VecTest and Rapid Analyte Measurement Platform (RAMP), and reverse transcriptase-polymerase chain reaction (RT-PCR) of oral swabs in a specialized viral transport medium (VTM). Results from this study demonstrated that VTM was excellent for transportation and maintenance of WNV in avian oral swab samples and allowed for detection by RT-PCR and subsequent confirmation by virus isolation. Oral swabs and kidney tissue in VTM tested by RT-PCR were found to have similar accuracy in detecting WNV in corvids. The two antigen-capture assays, VecTest and RAMP, provided few false positives for corvids, with over 95% specificity. When performed by multiple local agencies throughout the state, VecTest and RAMP were similarly sensitive for oral swabs of American Crows (Corvus brachyrhynchos) (70% and 64%, respectively). Data from known WNV positive corvid oral swabs in VTM tested by antigen-capture assays at a diagnostic laboratory suggested that RAMP was more sensitive than VecTest. Due to high probability of false negatives, neither test is recommended for use on non-corvids. While WNV antigen-capture assays were effective screening tools for corvids, they were markedly less sensitive for Western Scrub Jays (Aphelocoma californica).