Technical Abstract: Imported fire ant colonies were quantified in 1000 m2 circular subplots spaced about 125 m apart on a sheep and goat farm in Oklahoma. Mound counts and cumulative above-ground mound volume were used as measures of fire ant population density, and were subjected to regression analyses to determine effects of selected landscape metrics and habitat characteristics. The most important predictive variables correlated with fire ant population density were canopy cover (negative association) and incoming solar radiation (insolation, w*h/m2) at various times of year (positive association). Several predictive variables had small but significant effects on fire ant density, including distance from water, median silt content of soil, elevation, and distance from trees. A model predicting cumulative mound volume/plot with summer solstice insolation, elevation, distance from a cistern, distance from water, canopy cover, and distance from trees was the best and explained about 42% of the variation in population density. Results indicate that remotely sensed data in combination with publicly-available U. S. Geological Survey data may be useful in predicting areas of high fire ant abundance across areas at a field scale.