We compared two types of light traps used for monitoring mosquito abundance in the hydroelectric reservoir of Simplício, Além Paraíba - Minas Gerais. Mosquitoes were captured bimonthly using automatic CDC and Shannon traps before the filling of the hydroelectric plant reservoir from December 2008 to December 2009. In total, 1474 specimens from 13 genera were captured. Among the captured specimens, several species known to be vectors of disease-causing agents for humans and/or animals were identified, including Anopheles aquasalis, Aedes albopictus, Coquillettidia venezuelensis, Haemagogus leucocelaenus, and Aedes scapularis. Sampling efficacy between the four capture sites was not found to be significantly different, irrespective of species captured or type of trap used. Poor correlation (r (x, y) = -0.0444) between the number of mosquito species and capture site was observed when not influenced by the type of trap used. Among the installation sites of the CDC and Shannon traps in the areas investigated, CDC traps fixed in livestock shelters obtained an overall higher abundance of species captured.