Abstract Presenting with dismal prognoses, treatment of multifocal GBM (mGBM) requires extensive medical infrastructure. Individualized assessment for patients at risk of extended length of stay (LOS) and nonroutine discharge disposition, described as high-value healthcare outcomes, has not previously been described. To develop predictive models for extended LOS and nonroutine discharge disposition for mGBM patients. All mGBM patients undergoing surgical resection at our institution (January 1, 2009-December 31, 2020) were included. Predictive models for nonroutine discharge disposition (discharge to a non-home environment) and extended LOS [upper quartile of the LOS for all patients included in the present study’s cohorts (>6 days)] were developed using multivariate logistic regression models. Optimism-corrected C-statistics for each model were calculated using 1000 bootstrapped samples in order to correct for potential bias associated with the limited population size. The Hosmer-Lemeshow test was used to evaluate model goodness of fit and calibration. A total of 137 mGBM patients were included in the present study. Most patients were male (40.1%), white (83.2%), not Hispanic/Latino (97.8%), not married (55.5%) and underwent elective surgery (85.4%). The average age among our patient cohort was 55.41 ± 13.17, the average KPS score was 77.23 ± 14.54 and the average mFI-5 was 0.613 ± 0.760. On bivariate analysis, older age (p=0.015) and decreasing Karnofsky Performance Status (KPS) score (p<0.001) were significantly associated with nonroutine discharge disposition while emergency admission (p=0.010) and increasing frailty (mFI-5) score (p=0.003) were significantly associated with extended LOS. On multivariate analysis, increasing age (p=0.031) and decreasing KPS (p<0.001) predicted nonroutine discharge while emergent admission (p=0.023) and greater mFI-5 (p=0.044) predicted extended LOS. Our models for nonroutine discharge and extended LOS had optimism-correct c-statistics of 0.7507 and 0.7309 respectively and displayed adequate calibration. Following external validation, our models may have utility as a risk stratification tool for this multifocal GBM patient populations.