Knowledge of vector distribution is important for the design of effective local malaria control programs. Here we apply ecological niche modeling to analyze and predict the distributions of malaria vectors based on entomological collection points in the State of Roraima in the northern Brazilian Amazon Basin. Anopheline collections were conducted from 1999 to 2003 at 76 localities, all with active malaria transmission. A total of 13 anopheline species was identified from 17, 074 adult females collected: Anopheles darlingi, An. albitarsis s.l., An. nuneztovari, An. triannulatus s.l., An. braziliensis, An. peryassui, An. oswaldoi s.l., An. mattogrossensis, An. strodei, An. evansae, An. squamifemur, An. mediopunctatus s.l., An. intermedius. Anopheles darlingi, and An. albitarsis were the most frequently found species. An. squamifemur was found for the first time in Roraima. A distributional prediction model (genetic algorithm for rule-set prediction-GARP) and environmental variables were used to predicted potential distribution range for six anopheline species that occurred at ≥ 19 collection points. The method allows for the application of moderate sample sizes to produce distribution maps of vector species that could be used to maximize efficiency of surveys and optimize use of economic resources in epidemiology and control.