Abstract: To address the differentiated requirements of multiple contracts, such as the West-East Power Transmission Framework Agreement, bilateral negotiations, and centralized bidding, proposing a more practical and implementable medium- and long-term contract decomposition method that considers uncertainties like runoff and electricity prices for cascaded hydropower is currently a key research focus and a challenge to be resolved. This study introduces a medium- and long-term contract power generation decomposition method for cascade hydropower stations, tailored to different trading modes, based on the provincial power market structure with a high proportion of hydropower in the southwest region. By employing big data mining combined with scenario clustering, hourly electricity consumption characteristic curves for bilateral negotiation contracts across various industries are derived, and multivariate Gaussian distribution is used to model electricity price scenarios. Subsequently, a multi-objective optimization model is established to maximize the return on medium- and long-term contracts for difference, minimize contract performance deviation, and maximize energy storage at the end of the dispatch period. The model is refined to achieve a 720 h monthly simulation schedule, yielding hourly contract decomposition characteristic curves with the highest contract transaction proportion for each trading mode. Finally, using actual data from cascade hydropower stations in a specific basin as a case study, the results demonstrate that the proposed method's contract decomposition curves align with the differentiated trading characteristics of various modes, are more executable in the short term, and further enhance the marketization benefits of hydropower.