Two trapping methods were compared during a survey of the distribution of tsetse flies in the Mbororo cattle breeding area of the Central African Republic: (a) several traps dispersed throughout the riverine forest galleries and remaining only one day at each site: (b) one sentinel trap placed at the cattle drinking point and remaining for several days. The latter method was more reliable and is therefore recommended. The concentration of tsetse flies at the drinking points was negligible during the rainy season.