This study examines the hydrologic and climatic conditions that precede major flood events on Lake Ontario, with the purpose of understanding the potential for seasonal forecasts to inform lake level management. Seven late spring/early summer flood events are identified since 1949, including the record‐breaking flood of 2017. The surface climate, atmospheric circulation, and antecedent lake levels for the preceding winter and spring seasons are examined. Results suggest that flood events are caused by different combinations of high, initial wintertime water levels across all of the Great Lakes, anomalously wet winters across the entire Great Lakes basin, and wet spring conditions, particularly in the eastern part of the basin. Wet winters that precede flood events are often associated with La Niña conditions, while wet springs are often associated with a westward shift of the North Atlantic Subtropical High. As the critical drawdown period for Lake Ontario occurs in the fall, before the onset of anomalous winter or spring/summer inputs, a generalized additive model was used to predict April–August maximum monthly average Lake Ontario water levels using November levels for all Great Lakes, a nonlinear response to the wintertime Niño 3.4 index, and scenarios of April–May overbasin precipitation. The Niño 3.4 index significantly improves lake level predictions, suggesting that an El Niño‐Southern Oscillation signal may be useful for lake level management. Future work needed to verify the use of El Niño‐Southern Oscillation for Lake Ontario flood forecasting and to link the North Atlantic Subtropical High to predictions of springtime Great Lakes climate is discussed.