Abstract INTRODUCTION West Nile virus (WNV) immunoglobulin M (IgM) antibodies have been shown to persist for up to 500 days in certain patients. To evaluate the usefulness of immunoglobulin G (IgG) avidity assessment in the diagnosis of WNV infection, we analyzed 54 WNV IgM- and/or IgG-positive serum samples from 39 patients with neuroinvasive disease and 15 asymptomatic cases tested during a seroprevalence investigation. METHODS Serological tests (WNV IgM/IgG antibody detection, IgG avidity) were performed using commercially available enzyme-linked immunosorbent assays. RESULTS WNV IgM antibodies were detected in 47 (87%) samples. Acute/recent WNV infection was confirmed based on low/borderline avidity index (AI) in 44 IgM-positive samples (93.6%). In three IgM-positive samples (6.4%), high IgG AIs were detected, thus indicating persisting IgM antibodies from previous infections. All IgM-negative samples showed high AIs. Patients with WNV neuroinvasive disease tested within 30 days showed low AIs. In six patients tested 34-50 days after disease onset, AI was borderline (42%-60%), suggesting earlier WNV IgG maturation. Samples with the highest IgM values were associated with the lowest AIs (Spearman's rho coefficient -0.767, p < 0.001). CONCLUSIONS Our results indicate that IgG avidity differentiates current/recent WNV infection from persistent IgM seropositivity from the previous WNV transmission season both in patients with WNV neuroinvasive disease and in asymptomatic persons. A strong negative correlation between IgM antibody levels and AI indicates that in cases with very high IgM levels, determination of IgG avidity may not be necessary. As many patients showed rapid avidity maturation, low IgG avidity is indicative of WNV infection within the previous month.