In the summer of 1999, West Nile virus was recognised in the western hemisphere for the first time when it caused an epidemic of encephalitis and meningitis in the metropolitan area of New York City, NY, USA. Intensive hospital-based surveillance identified 59 cases, including seven deaths in the region. We did a household-based seroepidemiological survey to assess more clearly the public-health impact of the epidemic, its range of illness, and risk factors associated with infection. We used cluster sampling to select a representative sample of households in an area of about 7.3 km(2) at the outbreak epicentre. All individuals aged 5 years or older were eligible for interviews and phlebotomy. Serum samples were tested for IgM and IgG antibodies specific for West Nile virus. 677 individuals from 459 households participated. 19 were seropositive (weighted seroprevalence 2.6% [95% CI 1.2-4.1). Six (32%) of the seropositive individuals reported a recent febrile illness compared with 70 of 648 (11%) seronegative participants (difference 21% [0-47]). A febrile syndrome with fatigue, headache, myalgia, and arthralgia was highly associated with seropositivity (prevalence ratio 7.4 [1.5-36.6]). By extrapolation from the 59 diagnosed meningoencephalitis cases, we conservatively estimated that the New York outbreak consisted of 8200 (range 3500-13000) West Nile viral infections, including about 1700 febrile infections. During the 1999 West Nile virus outbreak, thousands of symptomless and symptomatic West Nile viral infections probably occurred, with fewer than 1% resulting in severe neurological disease.