Because riparian ecosystems are highly valued for their diverse ecological services, past and ongoing disturbances in riparian zones have led to extensive restoration efforts, litigation, and compliance monitoring of the effects of livestock grazing. Better understanding of the factors that influence cattle riparian use, especially in landscapes supporting threatened or endangered fish, could lead to improved predictions of management outcomes and riparian recovery for sustainable grazing systems. Although published models predict habitat selection by cattle, there is a gap in our understanding of cattle use, or occupancy, in riparian zones. As part of a long-term, multi-disciplinary project in a semi-arid riparian system in Oregon, USA, we collected 4 yr (2017-2020) of cattle telemetry data to identify factors affecting riparian use by cattle. We used beta regression in a Bayesian hierarchical framework to model the daily proportion of cattle locations in the riparian zone. We hypothesized that riparian use would 1) increase with increasing Julian date, temperature, solar radiation, days in pasture, and days since herding, and 2) decrease with higher humidity and precipitation. The best model predicted that use was greater with increasing days since herding, number of days grazing in a pasture, and Julian date, and lower as relative humidity increased. Daily riparian use by cattle averaged 0.167 (SD = 0.180) across years and pastures. The final model performed well, based on k-fold cross validation (Pearson's correlation = 0.72; 90% CI from 0.66 to 0.77). Our findings demonstrate the importance of considering management strategies (herding, grazing seasons) that affect riparian use by cattle, in tandem with weather, pasture characteristics, and other factors, and can be used in decision support systems to guide riparian grazing management.