The discrimination of methanol from chemically similar volatile organic compounds remains a challenge for chemiresistive gas sensors due to limited intrinsic selectivity. In this work, we investigate an electronic-nose approach based on a Ti₃C₂Tₓ MXene sensor array combined with machine-learning–based pattern recognition. The MXene sensors show reproducible responses to methanol, ethanol, and acetone at room temperature, and their transient response profiles enable the extraction of informative features for classification. Using supervised learning models, methanol can be distinguished from the other vapors with classification accuracies of up to 99% for measurements acquired on the same day. When applied to data collected several days later, the classification performance remains at approximately 83%, suggesting a degree of robustness against sensor drift. While further optimization and long-term studies are required, these results indicate that MXene-based sensor arrays coupled with data-driven analysis represent a promising direction for electronic-nose systems targeting volatile organic compounds discrimination.