Harmful algal blooms (HABs) in large rivers are increasing in both occurrence and intensity worldwide. Despite the immense data and understanding of HAB development in lentic systems, predicting and managing HABs in flowing waters can be difficult. For example, river plankton are often exposed to constantly changing light, temperature, and nutrient conditions due to continuous vertical transport from currents, tributary and point-source inputs, and dam-induced water velocity changes. We used seasonal laboratory bioassays to determine how light and temperature influence phytoplankton community composition, productivity, and maximum biomass in the 8th order Ohio River, USA. We also put these physical controls in context of nutrient/stochiometric regulation. One important finding was that different toxin producing cyanobacteria genera dominated at both ends of light and temperature ranges, with Microcystis sp. dominant above 750 μmol quanta m-2 s-1 and above 27 °C, and Dolichospermum sp. dominant below 150 μmol quanta m-2 s-1 and 18 °C. However, cyanobacteria dominance only occurred after the majority of available nutrients were sequestered, and only one of these genera can fix nitrogen. Results from this work are currently being used to parameterize a system-level hydrodynamic HAB prediction model of the a 99.4km long Greenup Pool of the Ohio River to better understand spatial and temporal HAB development, and potential mitigation strategies in large rivers.