OBJECTIVE: Although the use of osimertinib can significantly improve the survival time of lung adenocarcinoma (LUAD) patients with epithelial growth factor receptor mutation, eventually drug resistance will limit the survival benefit of most patients. This study aimed to develop a novel prognostic predictive signature based on genes associated with osimertinib resistance. METHODS: The differentially expressed genes (DEGs) associated with osimertinib resistance in LUAD were screened from Gene Expression Omnibus datasets and The Cancer Genome Atlas datasets. Multivariate cox regression was used to establish a prognostic signature, and then a nomogram was developed to predict the survival probability of LUAD patients. We used ROC curve and DCA curve to evaluate its clinical prediction accuracy and net benefit. In addition, the differentially expressed genes significantly associated with prognosis were selected for immune infiltration analysis and drug sensitivity analysis, and their roles in the progression of lung adenocarcinoma were verified by in vitro experiments. RESULTS: Our evaluation results indicated that the new nomogram had higher clinical prediction accuracy and net benefit value than the TN nomogram. Further analysis showed that patients with low STRIP2 expression had a higher level of immune response, and may be more likely to benefit from immune checkpoint inhibitors and conventional antitumor drugs. This may help to select more precise and appropriate therapy for LUAD patients with osimertinib resistance. Furthermore, in vitro experiments showed that STRIP2 promoted the LUAD cells proliferation, migration and invasion. This further demonstrates the importance of this gene signature for prognostic prediction. CONCLUSION: We developed a reliable prognostic model based on DEGs associated with osimertinib resistance and screened for biomarker that can predict the immune response in LUAD patients, which may help in the selection of treatment regimens after osimertinib resistance.