Abstract: OBJECTIVE To mine and evaluate the postmarketing safety signals of lopinavir/ritonavir recommended for trial in the novel coronavirus pneumonia(COVID-19) treatment regimen, and to provide references for rational drug use in the clinic. METHODS The reporting odds ratio and Bayesian confidence interval progressive neural network(BCPNN) method were used to detect the risk signals from the data in the adverse event reporting system database(FAERS) of the US Food and Drug Administration(FDA) with lopinavir/ritonavir as the primary suspected object from January 1, 2004 to December 31, 2019. The safety signals of gastrointestinal, liver, kidney, nervous and metabolic systems were focused on. RESULTS Of the 11 170 959 ADEs reports in the analysis, 10 120 ADEs were reported with lopinavir/ritonavir as the primary suspected drug, which involved in multiple system. High risk signals of clinical significance included acute pancreatitis(ROR=4.32, IC-2SD=1.65), cytolytic hepatitis(ROR=20.90, IC-2SD=3.66), hypertriglyceridemia(ROR=27.80, IC-2SD=4.07), cerebral ventricular dilatation(ROR=58.04, IC-2SD=4.26), acquired adipose dystrophy(ROR=165.80, IC-2SD=6.10), and so on. Other high risk safety signals not included in the drug specification were tachypnoea(ROR=5.27, IC-2SD=1.79), acquired Fanconi syndrome(ROR=122.34, IC-2SD=5.48) and mitochondrial toxicity(ROR=225.12, IC-2SD=5.61). These three safety signals were further confirmed by the time scan map that showed a strong statistical association with lopinavir/ritonavir. CONCLUSION The FAERS-based detection of ADEs signals show that close attention should not only paid to these ADEs such as acute pancreatitis, cytolytic hepatitis, hypertriglyceridemia, ventricular dilatation, and acquired adipose dystrophy, but also to the risks such as possible tachypnoea, acquired Fanconi syndrome, and mitochondrial toxicity and so on if lopinavir/ritonavir was used in the treatment of COVID-19.