Technical Abstract: Mosquito and West Nile virus surveillance was conducted on a National Wildlife Refuge in northeast Montana, 2005-2006, during which outbreaks of WNV in a colony of American white pelicans (Pelecanus erythrorhynchos Gmelin) resulted in juvenile mortality rates of 30 and 31%. During both years, floodwater species Ochlerotatus dorsalis (Meigen), Aedes vexans (Meigen), and Oc. flavescens (Muller) comprised 78% of the total collection with most host activity detected between mid-June and mid-July. Culex tarsalis Coquillett, the primary vector of WNV in this region, was most active between mid-July and mid-August and comprised 18% of the total collection in 2005 and 20% in 2006. However, 91% of the Cx. tarsalis females collected in 2006 were obtained from two traps located approximately 10 km away from the pelican’s nesting grounds. Culex tarsalis females that were collected closer to the roosting site comprised 1.7% of the total number of mosquitoes collected in 2006. Seasonal minimum infection rates per 1,000 Cx. tarsalis tested for WNV was 1.36 in 2005 and 1.41 in 2006. All pools in which WNV was detected in 2006 were composed of females collected from the two traps located 10 km from the nesting grounds. Substantial juvenile pelican mortality in 2006 despite the reduced population of the primary vector and mosquito infection rates near the colony suggests that the methods used to detect the introduction of WNV were too coarse and that amplification of the virus within the colony may have been due to routes other than mosquito infection.