To overcome challenges with traditional response assessment in anti-angiogenc agents, the current study uses T1 subtraction maps to quantify volumetric radiographic response in cabozantinib monotherapy, an orally bioavailable tyrosine kinase inhibitor with activity against VEGFR2, MET, and AXL, in an open-label, phase II trial in patients with recurrent glioblastoma (NCT00704288).A total of 108 patients with adequate imaging data and confirmed recurrent GBM were included in this retrospective study from a phase II multicenter trial of cabozantinib monotherapy (XL184-201) at either 100mg (N=87) or 140mg (N=21) per day. Contrast enhanced T1-weighted digital subtraction maps were used to define volume of contrast enhancing tumor at baseline and subsequent follow-up time points. Volumetric radiographic response (>65% reduction in contrast enhancing tumor volume from pre-treatment baseline tumor volume sustained for more than 4 weeks) was tested as an independent predictor of overall survival (OS).Volumetric response rate (VRR) for all therapeutic doses was 38.9% (41.4% and 28.6% for 100mg and 140mg doses, respectively). A log-linear association between baseline tumor volume and OS (P=0.0006) and a linear correlation between initial change in tumor volume and OS (P=0.0256) were observed. A significant difference in OS was observed between responders (median OS=20.6 months) and non-responders (median OS=8.0 months) (HR=0.3050, P<0.0001). Multivariable analyses showed continuous measures of baseline tumor volume (HR=1.0233, P<0.0001) and volumetric response (HR=0.2240, P<0.0001) were independent predictors of OS.T1 subtraction maps provide value in determining response in recurrent GBM treated with cabozantinib and correlated with survival benefit.