Abstract: Fractured-vuggy carbonate reservoirs of the Permian Maokou Formation in the central Sichuan Basin are highly heterogeneous, and conventional log responses of gas, water, and gas–water layers overlap to varying degrees, making fluid identification highly uncertain. To reduce this ambiguity, this study proposes a fluid identification scheme by combining principal component analysis (PCA) and Fisher discriminant analysis (FDA). Eleven fluid-sensitive log parameters are used to build a feature matrix, from which PCA extracts the first three principal components with a cumulative contribution rate of 82.6%. These components are interpreted in terms of reservoir-space development, combined variations in water responses and conductivity characteristics, and fluid occurrence. The principal components are then used as FDA inputs to construct classification functions for gas, gas–water, and water layers. Posterior probabilities are further used to define determinate and ambiguous discrimination zones. Based on 1,209 calibrated samples from 14 appraisal wells, the model yields an overall identification accuracy of 92.6%, with 99.1% for gas layers, 90.6% for water layers, and 72.6% for gas–water layers. Applications to typical wells show that the method achieves relatively stable results for pure gas and water layers. For gas–water layers that may be misclassified by conventional interpretation under a tight-matrix high-resistivity background, it also provides supplementary discrimination. The results can support log-based fluid evaluation of fractured-vuggy carbonate reservoirs of the Maokou Formation in the central Sichuan Basin. However, for complex intervals in gas–water transition zones or those with strong log-response overlap, oil testing and other logging data are still needed for integrated interpretation.