A longitudinal epidemiological survey of bovine trypanosomosis and its vectors was carried out in the Volta river basin of Northern Ghana to determine the relationship between cattle management and the incidence of bovine trypanosomosis. Two groups of sentinel cattle under different systems of management, classified as "fully-sedentary" and "partially-sedentary" (depending on the type of management) were followed over a 1-year period starting from March 2003 onwards. Cattle were screened at intervals of 3 months using the buffy coat technique (BCT). Buffy coat specimen from animals that were positive for the BCT and those that were negative, but with a packed cell volume (PCV) of less than 21% were further tested using the polymerase chain reaction (PCR). Plasma from all animals were tested for antibody using the indirect antibody enzyme-linked immunosorbent assay (ELISA). Trypanosomosis challenge was determined in tandem with the epidemiological survey with watering sites of sentinel cattle being the foci of interest. The parasitological prevalence at the start of the survey was higher in the fully-sedentary group (9%) than in the partially-sedentary group (3%). In subsequent visits, however, the parasitological incidence was consistently higher in the partially-sedentary group than in the fully-sedentary group. The mean seroprevalence (ELISA) of both groups increased from 3% in March to 54% in December. Statistical analysis of the serological results using a random effect logistic regression, showed a significant difference in incidence of bovine trypanosomosis between the two groups. There was also a significant effect of time. The influence of cattle herding on host-vector-parasite interface and its consequence on the incidence of trypanosomosis are discussed. © 2004 Elsevier B.V. All rights reserved.