In the Ghibe valley, south-west Ethiopia, a tsetse control trial using deltamethrin-impregnated targets was started in May 1990. The mean relative density of the main vector, Glossina pallidipes, fell from 2.1 flies per trap per day in the 12 months prior to introduction of tsetse control to 0.41 flies per trap per day in the 12 months after tsetse control was initiated. The annual mean prevalence of Trypanosoma congolense infections in cattle fell from 3.2 percent in the 2 years before tsetse control to 13 percent in the 2 years following deployment of targets. The largest reduction occurred in the first quarter of 1991 when the mean monthly trypanosome prevalence was 5 percent. There was also a decrease of approximately 75 percent in the estimated rate of relapse of trypanosome infections in cattle after treatment with diminazene aceturate. However, in June 1991, socio-political disturbances occurred in Ethiopia and were associated with substantial thefts of targets. Following these thefts, the mean relative density of G. pallidipes rose from 0.41 flies per trap per day in the period May 1990-April 1991, to 1.29 flies per trap per day in the period April 1992-March 1993. Associated with this rise, the mean trypanosome prevalence in cattle increased from 13 percent in the period May 1991- April 1992 to 28 percent in the period May 1992-April 1993. This was similar to the mean trypanosome prevalence in cattle during the pre-control period. Thus, while effective tsetse control methods can be used to reduce the transmission of trypanosomes their long- term impact is dependent on their sustainability.