Objective To construct a prognostic model for the overall survival of intracranial germ cell tumors （iGCT） based on the Surveillance， Epidemiology， and End Results （SEER） database， so as to predict prognosis and guide treatment. Methods Clinical pathological and sociodemographic data of 1 105 iGCT patients from 1975 to 2020 were collected from the SEER database. The patients were randomly divided into a training set and a validation set in a 1∶1 ratio. Univariate and multivariate Cox regression were used to screen for independent prognostic factors and construct the Outcome Scale to predict the overall survival of iGCT patients. The area under the receiver operating characteristic curve （AUC）， consistency index （C-index）， calibration curve， and decision curve analysis （DCA） indicators were used to evaluate the performance of the Outcome Scale. Results The median follow-up time for all iGCT patients was 84 months， including 79 months for adult patients and 87 months for pediatric patients. Primary site ［HR=0.616， 95%CI （0.451， 0.843）， P=0.002］， radiotherapy ［HR=0.390， 95%CI （0.286， 0.533）， P＜ 0.001］， chemotherapy ［HR=0.656， 95%CI （0.475， 0.905）， P=0.01］， and age ［HR=0.658， 95%CI （0.479， 0.905）， P=0.01］ were independent prognostic factors for overall survival， with statistically significant differences. The Outcome Scale for predicting overall survival constructed using these clinical factors showed that the AUC of the 1-year， 5-year， and 10-year survival rates in the training set were 0.751， 0.658， and 0.644， respectively， and the C-indices were 0.739， 0.645， and 0.649， respectively. The AUC of the 1-year， 5-year， and 10-year survival rates in the validation set were 0.862， 0.752， and 0.734， respectively， and the C-indices were 0.850， 0.716， and 0.720， respectively. The calibration curve had a good overlap with the standard reference curve. DCA suggested that the training set of the model had a high net return between risk probabilities of 0.125 and 0.25， while the validation set has a high net return between risk probabilities of 0.25 and 0.375. The model had a good overlap with the curve of the full variable prediction model. Conclusions The primary site， treatment method， age， and prognosis of iGCT patients are correlated， and the Outcome Scale constructed based on this can help predict patient prognosis and guide treatment.