Long-term survival is an often used, yet poorly defined, concept in the study of glioblastoma multiforme (GBM). This study suggests a method to define a time-point for long-term survival in patients with GBM. Data for this study were obtained from the Surveillance, Epidemiology and End-Results database, which was limited to the most recent data using the period approach. Relative survival measures were used and modelled using piecewise constant hazards to describe the survival profile of long-term survivors of GBM. For patients with GBM, the first quarter of the second year (5th quarter) post-diagnosis is considered to be the peak incidence of mortality with an excess hazard ratio of 7.58 (95% confidence interval=6.54, 8.78) and the risk of death due to GBM decreases to half of its rate at 2.5 years post-diagnosis. The 2.5-year cumulative relative survival (CRS) for all patients is approximately 8%, with a CRS of approximately 2% at 10 years. Using the definition of long-term survival suggested here, the results indicate that long-term survivors are patients who survive at least 2.5 years post-diagnosis. The most likely time period for patients with GBM to die is the 5th quarter post-diagnosis