The sampling efficiency of light trap catches relative to human bait catches in estimating biting rates of the mosquito Anopheles gambiae Giles was investigated in two types of community in southern Sierra Leone: (i) where most of the inhabitants slept under treated bed nets; and (ii) where most of the inhabitants slept without bed nets. The number of female A. gambiae mosquitoes caught in these communities by light trap was strongly correlated (r > or = 0.72) with those from corresponding human biting catches performed either on the same or adjacent nights. It was found that the relative sampling efficiency of light traps varied slightly but significantly with mosquito abundance in villages with treated bed nets, but not in those without them. Nevertheless, the relationship between relative sampling efficiency and mosquito abundance did not differ significantly between the two types of village. Overall, there was insufficient evidence to show that the presence of treated nets altered the relative efficiency of light traps and any bias was only slight, and unlikely to be of any practical importance. Hence, it was concluded that light traps can be used as a surrogate for human bait catches in estimating biting rates of A. gambiae mosquitoes in the two communities.