A West Nile virus (WNV) surveillance system was created and implemented in New York State (NYS) in 2000 and described previously (White et al. 2001). We examine and evaluate the results of mosquito and virus surveillance for 2000 through 2004 exclusive of New York City. Forty-nine counties submitted 1,095,426 mosquitoes in 35,280 pools for WNV assay. Specimens of 47 species were tested, with Culex species accounting for 47.6% of all pools tested. WNV was detected in 814 pools from 10 species, with combined Culex pipiens/Culex restuans pools accounting for 90.8% of all detections. Pools submitted from gravid traps were 5.7 times more likely to be positive than submissions from carbon dioxide-baited light traps. Most human WNV cases resided in counties that conducted mosquito surveillance. Local health departments' use of mosquito surveillance information often led to an enhanced disease prevention response. In NYS, Cx. pipiens/Cx. restuans groups are most likely vectors of WNV. Future efforts to improve system efficacy are discussed.