Abstract: With the continuous improvement of exploration degree in Pearl River Mouth Basin, drilling complexities and operational risks have been progressively increasing. Oil-based mud has been widely applied in exploration to ensure drilling safety and protect reservoirs. However, owing to the principle of "like dissolves like", the white oil in oil-based mud significantly interferes with gas logging components during mud logging, which severely restricts the rapid identification of fluid types while drilling. It is therefore imperative to develop a reliable identification model for fluid types under oil-based mud conditions. By comparing the actual gas logging data of two wells, this paper deeply analyzes the influence of oil-based mud on gas logging responses. Based on the differences in gas logging components between water-based mud and oil-based mud systems from these wells, a set of correction coefficient F_n for the C₁～C₅ components is derived, and an accurate identification method for fluid types using oil-based mud gas-logging data is thereby established. This method has been successfully applied to real-time fluid-type identification while drilling in Huizhou Sag, which effectively corrects the interference caused by oil-based mud, significantly improves the coincidence rate of mud logging interpretation, and achieves favorable application effects. This study provides a feasible approach to solve the difficulty of fluid type identification under oil-based mud systems, and holds good potential for broader application.