Determination of infection incidence and prevalence is essential in the epidemiological study of cattle trypanosomoses. These studies are based on direct and indirect diagnostic techniques. Parasitological techniques are of low sensitivity. PCR detection has a better sensitivity, which is not, however, sufficient to detect most chronic infections. Detection by indirect-ELISA of antibodies directed against trypanosomes is the best-adapted tool to assess the magnitude of trypanosome infections in cattle populations. However, the detection of antibodies in cattle does not mean the animals are actively infected; determining antibody persistence after curative treatment or natural self cure are essential in analyzing the results. In the present paper, persistence of anti-trypanosome antibodies was evaluated by indirect-ELISA in 32 crossbred cattle (zebu x Baoule) naturally infected by Trypanosoma vivax, after treatment with diminazene aceturate (7 mg/kg i.m.). The initial seroprevalence was 100%; the decline in antibodies started three months later to reach a seroprevalence of 3% five months after treatment. Persistent positive cases were due to the persistence of the infection or to natural reinfection in two animals during the study (they were excluded from the cohort), and to one case unaccounted for. Seroprevalence decreased faster in the young. It is suggested that the seroprevalence by indirect-ELISA against T. vivax four or five months after a curative treatment indicates the prevalence of active infections.