The objective of this study was to apply transition models to distinguish between factors associated with both incident and persistent trypanosome infections. Data collected from 1,561 cattle were analyzed from a long-term study involving 8 herds in which both trypanosome infections (a total of 56931 cattle sampling-months) and tsetse (Glossina spp.) challenge were monitored monthly from March 1986 to March 1998. Both pour-on and insecticide-target tsetse control programs and mass treatment with diminazene aceturate before tsetse control were associated with significant decreases in both incidence and persistence of trypanosome infection relative to noncontrol periods, as were seasonal and sex effects. The magnitudes of the effects were, however, often different for new and persistent infections. For persistence of infection, there were 2 trends. In general, the duration of infection increased during the study, despite the regular treatment with diminazene aceturate. The transition model had 2 major benefits. The first was to identify an increasing duration of infections with time, taking into account other factors associated with increasing infection risk. The second was to highlight different patterns in the effects of certain factors on new and persistent trypanosome infections.