Predator–prey interactions are an inherently local-scale phenomenon, but the intensity of these interactions can be mediated by abiotic conditions that can exert a multi-scaled influence through space and time. Understanding how multi-scale abiotic factors may influence local-scale biotic processes has proven challenging; however, the hierarchical nature of riverine flow regimes makes these environments an ideal setting to test how predator–prey relationships may vary with multi-scaled flow variation. We developed a series of Bayesian hierarchical models to explore how predator–prey relationships between barramundi Lates calcarifer and their prey may be influenced by multi-scaled flow variables in the Daly River, northern Australia. We found that spatio-temporal variation in barramundi abundance was strongly related to both antecedent flow and the abundance of prey fishes (predictive r2 = 0.57), and that barramundi abundance is more likely to be influenced by bottom-up, rather than top-down predator–prey dynamics. We also found that the strength and direction of these relationships varied across the catchment and between seasons. We found stronger, positive relationships between barramundi abundance and prey abundance in the most downstream sites with higher mean annual flows, compared to upstream sites. These results indicate that the abundance of predatory fishes can be related to both recent abiotic (flow) conditions and the abundance of prey (biotic conditions), and provides strong support for the importance of bottom-up trophic dynamics. Management of iconic predators such as barramundi should therefore consider both flow management and other key factors such as habitat maintenance to support their prey.