Despite a control programme, malaria incidence in Yunnan is increasing. Population movements for social and cultural reason expose non-immunes to high transmission areas. Transmission also occurs within forest-fringe villages. The vectors Anopheles dirus and An. minimus are species complexes, exhibiting behavioural variations. Knowledge of vector bionomics is needed for efficient control. Multidrug-resistant Plasmodium falciparum necessitates alternatives to human landing catches. CDC light traps with UV or ordinary incandescent bulbs were used for 57 trap-nights: 2703 mosquitoes were caught, comprising the vector species An. minimus and An. sinensis and the supected vector An. maculatus. Larval An. dirus were found around the village bu no aduldts were trapped. UV light trpas caught more mosquitoes then the incandescent traps, but were non-specific and unpopular with villagers. Traps placed in living areas of houses caught more mosquitoes than those placed beside bednets, and the catch comprised more species that were active in the early evening. Evans' traps hung outdoors baited with CO2 caught few mosquitoes. CDC traps in the same position baited with CO2 caught large numbers of Culex tritaeniorhynchus. Indoor spray catches recovered human-fed An. vagus and An. minimus. This work confirmed that CDC light traps could be used to trap local vectors, and the abundance of early active mosquitoes in the living area suggests that personal projection measures may be required in the evening, to supplement bednet use.