Flood events are influenced by terrestrial factors including land cover, land and water management, watershed physiographic features, and hydro-climatic components including snowmelt and precipitation. In Canada, flooding is a frequent and prominent natural disaster, which is modulated by different flood-generating mechanisms. In this study, we assess the intensity and frequency of three flood-generating mechanisms including Rain on Snow (ROS), intense rainfall, and snowmelt-driven flood events over the Assiniboine-Red River basin, which is one of the most flood-prone regions in Canada and located in the Lake Winnipeg watershed. We downscale and bias correct seven Global Climate Models (GCMs) that participated in CMIP6 using two methods of Bias Correction/Constructed Analogues with Quantile mapping (BCCAQ) (BCCAQ) and Multivariate Bias Correction (MBC). The observed and downscaled climate variables (precipitation and temperature) are used to drive a process-based distributed snow model to evaluate the changes in flood-generation mechanisms in the historical and future periods. The projected future changes are analyzed under policy-relevant global mean temperature (GMT) increases from 1.0 °C to 3.0 °C above the pre-industrial period. Overall, all models project higher regional temperature increases compared to the global mean with warmer and wetter winters. The snow model results indicate future decreases in the snow cover duration, snowmelt rate, and snow water equivalent (SWE), and earlier shifts in the maximum SWE timing. Moreover, both the intensity and frequency of ROS events increase in all seasons except summers. However, the increases in the rain and snowmelt events are mostly projected to occur in the spring.