In recent years, domestic listed pharmaceutical enterprises frequent financial fraud incidents, to the society and investors have a great negative impact. How to effectively identify financial anomalies of pharmaceutical enterprises and avoid financial risks is the focus of academic and audit circles. First of all, based on the relevant literature of this study, the causes and characteristics of financial risks of pharmaceutical enterprises are sorted out, and the index system for financial risk identification of pharmaceutical enterprises is summarized, including six first-level indicators and 19 s-level financial indicators. Secondly, the financial risk identification model of pharmaceutical enterprises is constructed, and the financial data of 25 normal and 53 abnormal pharmaceutical enterprises from 2013 to 2022 are obtained according to CSMAR database. Finally, the characteristics of the selected indicators are described by clustering method, and then incorporated into the random forest model for statistical testing and evaluation of the index results. The research results show that among the many models that can identify the financial risks of pharmaceutical enterprises, the random forest model constructed in the paper has a good recognition effect, and the stable accuracy rate reaches 77% in capturing high-latitude financial data and output data, which is beneficial to lay a good foundation for the establishment of a suitable financial risk early warning mechanism for pharmaceutical enterprises. Finally, based on the results of random forest evaluation, the paper puts forward the prospect of preventing financial anomaly identification of pharmaceutical enterprises.