Modification of river flows is a major cause of freshwater fish population declines in many parts of the world. Identifying the precise mechanisms of these declines represents a significant challenge, as a range of stressors can simultaneously impact various components of fish health, fitness and population dynamics. Here we investigate the role of river flows and other biophysical factors on spatio-temporal variation in freshwater fish body condition in Australia's highly modified Murray-Darling Basin using three widely distributed native (Murray cod Maccullochella peelii, golden perch Macquaria ambigua and bony herring Nematalosa erebi) and one introduced (common carp Cyprinus carpio) species. Our aim was to uncover drivers of spatio-temporal variation in fish condition at two spatial extents: at the basin scale, utilising a flow regime disturbance index, and at the river-valley scale, employing individual flow gauge data to assess responses in fish condition to multiple measures of antecedent (365 day) flow. Linear mixed effects modelling revealed that at the basin scale, M. peelii and M. ambigua were in better condition in rivers with lower flow regime disturbance, and temporal trends in the condition of N. erebi, C. carpio and M. peelii reflected boom and bust dynamics related to wet and dry climate periods. At the river-valley scale, mean antecedent daily flow magnitude was significantly positively related to the condition of M. peelii, M. ambigua and C. carpio, whereas the number of high-flow days was negatively related to condition of N. erebi. Our study demonstrates that a simple body condition index calculated from routinely collected length-weight data is sensitive to multiple measures of hydrological disturbance in river systems that experience substantial temporal and spatial variability. We emphasise that studies considering multiple spatial scales are important for understanding complex scale-dependent mechanisms influencing fish condition.