Mosquito traps known as Mbita traps made from modified bednets according to a design developed in Kenya were compared with Centers for Disease Control (CDC) light traps for their ability to catch anopheline and culicine mosquitoes in several different villages in northern Tanzania. The results confirmed those recently reported in Kenya, that Mbita traps catch significantly fewer mosquitoes than CDC traps. Statistical analysis using a Poisson log linear model with random effects for the trap counts showed that the ratio of the catches with the two types of trap was not consistent in the different villages. Thus, we doubt whether the Mbita trap would be a reliable substitute for CDC traps. In one trial, the catches made at different hours of the night with the two types of trap indicated that in villages where insecticide treated nets (ITNs) had been used for some years, somewhat more of the Anopheles biting occurred early and late in the night, whereas in villages with no history of ITN use, biting was concentrated in the middle of the night. This suggests that behavioural adaptation to avoid contact with ITNs may be beginning to evolve.