An average of 840 East African Zebu cattle from nine herds in the Ghibe Valley, southwest Ethiopia were monitored from 1986 to 1990. Each month blood samples were collected from analysis of packed red cell volume and detection of trypanosomes. Animals were found to be parasitaemic and with a PCV less than 26 percent were treated with diminazene aceturate at a dose of 3.5 mg/kg body weight. The majority of infections were associated with trypanosoma congolense (84 percent of infections in adult cattle and 71 percent in cattle less than 24 months of age), and the mean percentage of adult animals detected parasitaemic 1 month after treatment of an infection with T. congolense was 27 percent. In order to assess possible existence of drug resistance, a model was applied which allowed monthly incidences of new infections to be distinguished from recurrent infections. This model showed the monthly incidence of new infections of T. congolense in adult cattle increased significantly from 11 percent in 1986 to 24 percent in 1989 following a concomitant increase in the tsetse challenge. The corresponding increase in overall prevalence of T. Congolense was from 17 percent to 38 percent and the mean prevalence of recurrent infections increased significantly from 6 percent to 14 percent. These findings ruled out the possibility that the high prevalence of trypanosome infections in cattle was due only to a high tsetse challenge and pointed to the existence of T. congolense populations which expressed resistance to diminazene. There were variations associated with season, herd, age and sex in the incidence of new infections, prevalence of recurrent infections and relapse to treatment.