The distribution of West Nile virus (WNV) is dependent on the occurrence of both susceptible avian reservoir hosts and competent mosquito vectors. Both factors can be influenced by geographic variables such as land use/landcover, elevation, human population density, physiographic region, and temperature. The current study uses geographic information systems (GIS) and logistic regression analyses to model the distribution of WNV in the state of Georgia based on a wild bird indicator system, and to identify human and environmental predictor variables that are important in the determination of WNV distribution. A database for Georgia was constructed that included (1) location points of all the avian samples tested for WNV, (2) local land use classifications, including temperature, physiographic divisions, land use/landcover, and elevation, (3) human demographic data from the U.S. Census, and (4) statistics summarizing land cover, elevation, and climate within a 1-km-radius landscape around each sample point. Logistic regression analysis was carried out using the serostatus of avian collection sites as the dependent variable. Temperature, housing density, urban/suburban land use, and mountain physiographic region were important variables in predicting the distribution of WNV in the state of Georgia. While weak, the positive correlation between WNV-antibody positive sites and the urban/suburban environment was consistent throughout the study period. The risks associated with WNV endemicity appear to be increased in urban/ suburban areas and decreased in the mountainous region of the state. This information may be used in addressing regional public health needs and mosquito control programs.