In the Chaganhua Gas Field within the Songnan Fault Depression, the Huoshiling Formation's volcanic reservoirs exhibit an average porosity of 4.5% and a permeability of 0.08×10^-3 μm, indicating a dense and highly heterogeneous nature. Due to this complexity, a comprehensive approach, testing a broad set of reservoirs, is required to establish effective classification criteria. This study used physical property data, high-pressure mercury injection, nuclear magnetic resonance, and other experiments to analyze the microstructure of volcanic reservoirs. Through multi parameter comparison, a microscopic classification standard was established. Nuclear magnetic logging served as a bridge between microscopic and macroscopic parameters, facilitating the creation of a comprehensive evaluation framework for classifying volcanic reservoirs. This framework encompasses microscopic structural features such as pore throat radius, displacement pressure, mercury saturation, alongside macroscopic parameters obtained from nuclear magnetic logging and other experiments, such as the T2 spectrum distribution, centrifugal saturation, porosity, permeability, saturation, acoustic time difference, lithology density, and resistivity. Reservoirs are categorized from high to low quality into classes A, B, and C based on this comprehensive set of criteria. This method has strong operability and provides a reliable basis for the testing plan of new drilling and the optimization of sweet spots in exploration and development of horizontal wells. The research methods and understanding have certain reference significance for the classification research of volcanic reservoirs.