Abstract: Objective To establish and validate a nomogram of glioblastoma prediction model to estimate individualized survival probability.Methods The Surveillance, Epidemiology and End Results(SEER) database of the National Cancer Institute was used to collect data on the clinical and tumor features of 5 417 glioblastoma patients from 2011 to 2015.A least absolute shrinkage and selection operator(LASSO) model and Multivariate COX regression analysis were adopted to screen the independent risk factor associated with prognosis for the prediction model. The prediction precision and discriminatory power of the nomogram were assessed using a receiver operating characteristic curve(ROC) and calibration curve.Results The prediction model was developed, using age, surgery, chemotherapy and radiotherapy as prognostic factors. The area under the ROC curve(AUC) was 0.757 for the training set and 0.754 for the validation set. The calibration curve for determining the probability of survival indicated that the model prediction results were in good agreement with the actual observations. According to LASSO-COX risk regression, age, surgery, chemotherapy and radiotherapy were independent risk factors, while gender, marriage, race, laterality, primary site, tumor size and the number of tumor were poorly correlated with patient survival.Conclusions The nomogram survival prediction model was successfully established, providing a concise and reliable survival prediction tool for patients with glioblastoma.