In the late summer and fall of 1999, New York State (NYS) had the first outbreak of West Nile (WN) virus encephalitis in the Western Hemisphere (1). The nucleotide sequence of the viruses isolated during this outbreak was most similar to that of a 1998 isolate from a goose in Israel (2). By the end of 1999, 62 human cases, 7 fatal, had occurred in New York City (NYC) and two neighboring counties, Nassau and Westchester (3). Although WN virus infection was confirmed in dead birds shortly before it was confirmed in humans, no WN viruspositive dead birds were identified from time periods before the onset of symptoms in the first human cases, despite subsequent WN virus testing of birds collected earlier (4). Whether dead bird surveillance could provide an early warning for human WN virus cases could not be definitively established by analyses of 1999 data on dead bird surveillance. However, sightings of dead crows preceded laboratory confirmation of viral activity in any species, and testing of dead birds provided valuable information about the temporal and geographic spread of the virus (4). We evaluate the usefulness of dead bird surveillance in 2000 for detecting geographic spread of WN virus and providing an early warning of the risk for transmission to humans. We also discuss lessons learned for other states that may be instituting a similar system. Methods For WN virus surveillance, the New York State Department of Health (NYSDOH) developed and implemented an integrated electronic system based on the department's existing infrastructure for secure web-based electronic health information interchange with local health units, health-care facilities, and providers (5). The functional component of the infrastructure is called the Health Information Network, into which local health units entered data about sightings of ill or dead birds. Freshly dead birds were submitted by local health units to the New York State Department of Environmental Conservation's Wildlife Pathology Unit for necropsy, which included evaluation of gross pathologic indications of WN virus infection and other possible causes of death. Organs collected for laboratory testing included brain, kidney, heart, liver, and spleen. Necropsy results were entered by the Wildlife Pathology Unit into the Health Information Network. Local health units were permitted to send any species of birds for possible necropsy and WN virus testing. However, American Crows, Blue Jays, and Fish Crows, members of the Corvid family, which was most affected by the WN virus outbreak in 1999, were a top priority for submission, followed by raptors and house sparrows. As the outbreak progressed, birds from counties without documented WN virus were given higher priority, as well as migrating species of birds. Most laboratory testing on dead birds was done at the NYSDOH Wadsworth Center, as described (6). WN virus infection was confirmed by at least two positive assays.