The number of tsetse visiting a black cylindrical model (designed to look like a small host animal) in the Zambesi Valley, Zimbabwe, was estimated by catching them on an expanse of electric netting placed next to the model. The number which alighted was found by electrocuting the flies on the surface of the model on other occasions when the netting was absent. About half the Glossina morsitans morsitans Westw. visiting a stationary small model (a black drum 57 cm long and 37 cm in diametter) in the absence of ox odour alighted on it, but virtually none of the G. pallidipes Aust. alighted. When the linear dimensions of the standard model were increased threefold, about 1·5 times as many tsetse visited its close vicinity. Of these, nearly all the G. morsitans and about one-third of the G. pallidipes alighted on the model. Four to seven times as many tsetse visited the models when the odour of one ox was dispensed near them. When odour was present, about 1·5 times as many tsetse visited the large model as visited the small one. Some 55–85% of G. morsitans visiting the small model in the presence of ox odour alighted, but only about 30% of the male and 20% of the female G. pallidipes alighted. The latter percentages increased to 53 and 62 when the large model was used. A smaller percentage of G. morsitans females landed on the large model in the presence of ox odour than in its absence. With mobile models, significantly more tsetse were taken if electric nets were appended fore and aft of the model. In the absence of electric nets, 2–30 times as many tsetse landed on the large model as on the small one, the effect being greater for female G. morsitans than for males and much greater for G. pallidipes than for G. morsitans. The percentage of visiting flies of both sexes and species that landed was greater when the large model was used.