Abstract: For developed gas reservoirs with variable condensate content in East China Sea, accurate assessment of condensate contamination is challenging under multi-layer commingling production due to limited PVT data. Therefore, a contamination diagnosis and risk classification method based on production dynamic data needs to be established to support the development strategy formulation of offshore gas fields.This paper proposes a dynamic data-based method for contamination diagnosis and risk classification. A three-level risk standard (high, medium, low) is defined. For data-scarce cases, a quantitative model linking maximum gas-oil ratio and retrograde liquid saturation enables contamination evaluation in the absence of PVT data. Field applications show the method effectively diagnoses contamination history, provides early warnings, and delivers reliable assessments even with insufficient PVT data. It offers a practical tool for the pollution assessment of offshore multi-layer condensate gas reservoir development, supporting decision-making for development countermeasures, and has good application value.