In this work, a novel and efficient determination method for the rapid identification of multi-class chemical constituents of Ermiao Wan has been developed. UPLC-Q-TOF-MS/MS fingerprinting was coupled to a data mining method to more efficiently detect ion signals from accurate mass data. Ermiao Wan (EW) is a combination formula of Cortex Phellodendri Chinensis and Rhizoma Atractylodis that is commonly used to treat gout and hyperuricemia, as described in the State Pharmacopoeia of the People's Republic of China. However, the chemical constituents of EW remain unclear to date. In the present study, the multiple compounds in EW were extracted and simultaneously determined under optimized conditions. A Waters UPLC BEH C₁₈ column was used to separate the target analytes, followed by tandem mass spectrometry (MS/MS) detection using an electro-spray ionization source in positive and negative mode. As a result, a total of 100 compounds (45 ions in positive mode, 55 ions in negative mode) were detected. Among them, 93 components were tentatively identified by analysing the retention times and subsequent mass spectrometry fragment ions. UPLC-Q-TOF-MS/MS analysis revealed the complexity of the chemical composition of this formula. The integration of the multivariate data mining (MMA) with UPLC-Q-TOF-MS/MS was demonstrated to be a valuable strategy for the rapid screening and identification of major EW constituents.