In order to scientifically identify the most stringent water management area types for the 16 cities of Yunnan Province, the optimalforaging algorithm (OFA)-Projection Pursuit (PP) recognition model was proposed. The index system and grading standards were establishedaccording to the most stringent water resources management including the total amount of water control red line, water efficiency control red line,and pollution control red line. The index data and grading standard threshold of the various cities of Yunnan Province in 2015 were used to buildthe projection index function, the OFA was used to search the optimal projection vector, and the comprehensive projection values and thethreshold values of the various cities were calculated, and the graded standard threshold projection values were used to identify the most stringentwater resource management types in each region. The results show that the OFA-PP identification model has a good recognition effect, and therecognition results can provide a reference for the implementation of the most stringent water resources management system in Yunnan Province.