The classic Dunne Diagram depicts various runoff generation mechanisms governed by climate, soil, and topographic conditions, illustrating their qualitative relationships under long-term average conditions. This study revisits the Dunne Diagram using a newly derived theoretical framework for runoff generation. This new framework mathematically unifies the infiltration and saturation excess runoff in a single runoff scheme that determines the runoff produced by both generation mechanisms as a function of the evolving soil moisture variations within a catchment. The framework depicts the transition between the two mechanisms in space and time. Leveraging the observational data at over 180 natural catchments in the U.S. under various climatic and landscape conditions, we revisit the Dunne Diagram in several aspects by: 1) developing a “data-driven”, quantitative version of the Dunne Diagram; 2) examining the relative dominance of different runoff mechanisms across multiple temporal scales; 3) exploring a continuous representation of the relative importance of various runoff mechanisms, highlighting the transitional zones. By improving the understanding and modeling of runoff at the catchment scale, our work aims to improve flood forecasting and water resources management, particularly for regions with complex or changing hydrologic regimes. Further, it provides a more nuanced and comprehensive view of runoff generation processes, bridging the gap between theoretical concepts and real-world observations.