Objective Two prognostic models consisting of autophagy-related gene （ATG） or autophagy-related lncRNA （ATL） were constructed to predict the prognosis of patients with glioblastoma multiforme （GBM） and provide the basis for personalized diagnosis and treatment.Methods The RNA sequencing and clinical data of GBM in TCGA database were used to screen the differentially expressed ATG and ATL with prognostic value by univariate Cox regression analysis. Then， the prognostic models were constructed by multivariate Cox regression analysis， the risk values of patients were calculated according to the models and their effectiveness was verified.Results Compared with normal tissues， 72 differentially expressed ATG （up-regulated 53， down-regulated 19） and 170 differentially expressed ATL （up-regulated 102； down-regulated 68） were obtained. 16 ATG and 22 ATL were identified by univariate Cox regression analysis and multivariate Cox regression analysis included 3 ATG and 8 ATL， respectively， to construct the prognosis model. The survival curves of the two models showed that the survival rate of the high-risk group was much lower than that of the low-risk group， and the receiver operating characteristic curves proved that both models had good predictive ability.Conclusions The two prognostic models can effectively predict the survival of GBM patients， and provide reference for personalized diagnosis and treatment. ［Citation:Journal of International Neurology and Neurosurgery, 2021, 48(4): 359-365.］