The vertical distribution of mosquitoes was investigated at Keneba in The Gambia, using flight traps and suction traps set up over open ground. Flight traps were used to sample three levels, ground to 1·37 m, 1·45–2·82 m, and 2·90–4·27 m. Suction traps were mounted in two series of experiments at three levels, 0·68, 2·13 and 3·51 m, and in a third series at four levels, 0·91, 305, 6·10 and 9·15 m. Most species, including Anopheles melas Theo., A. squamosus Theo., Aedes punctothoracis (Theo.), the Culex decens group, C. tritaeniorhynchus Giles, Mansonia africana (Theo.) and M. uniformis (Theo.), were most common near the ground, with catches declining progressively with height. C. thalassius Theo. was frequent at all levels, although the largest catches were taken at 9·15 m. Only C. neavei Theo. and C. weschei Edw. were more common in the higher traps. Although larger catches of A. melas and C. tritaeniorhynchus were taken during moonless periods, the numbers of other species were unaffected by lunar phase. In general, the presence or absence of moonlight had little effect on vertical distribution. However, C. thalassius and the C. decens group were shown to be more attracted to flight traps in moonlight, and it is concluded that visual responses were involved. A limited series of four-level, human-bait catches, with the catchers stationed at the same heights as the suction traps of the third series, produced results similar to those of the suction trap collections.