Purpose To determine whether regions of low apparent diffusion coefficient (ADC) with high relative cerebral blood volume (rCBV) represented elevated choline (Cho)-to-N-acetylaspartate (NAA) ratio (hereafter, Cho/NAA ratio) and whether their volumes correlated with progression-free survival (PFS) and overall survival (OS) in patients with glioblastoma (GBM). Materials and Methods This retrospective analysis was approved by the local research ethics committee. Volumetric analysis of imaging data from 43 patients with histologically confirmed GBM was performed. Patients underwent preoperative 3-T magnetic resonance imaging with conventional, diffusion-weighted, perfusion-weighted, and spectroscopic sequences. Patients underwent subsequent surgery with adjuvant chemotherapy and radiation therapy. Overlapping low-ADC and high-rCBV regions of interest (ROIs) (hereafter, ADC-rCBV ROIs) were generated in contrast-enhancing and nonenhancing regions. Cho/NAA ratio in ADC-rCBV ROIs was compared with that in control regions by using analysis of variance. All resulting ROI volumes were correlated with patient survival by using multivariate Cox regression. Results ADC-rCBV ROIs within contrast-enhancing and nonenhancing regions showed elevated Cho/NAA ratios, which were significantly higher than those in other abnormal tumor regions (P < .001 and P = .008 for contrast-enhancing and nonenhancing regions, respectively) and in normal-appearing white matter (P < .001 for both contrast-enhancing and nonenhancing regions). After Cox regression analysis controlling for age, tumor size, resection extent, O-6-methylguanine-DNA methyltransferase-methylation, and isocitrate dehydrogenase mutation status, the proportional volume of ADC-rCBV ROIs in nonenhancing regions significantly contributed to multivariate models of OS (hazard ratio, 1.132; P = .026) and PFS (hazard ratio, 1.454; P = .017). Conclusion Volumetric analysis of ADC-rCBV ROIs in nonenhancing regions of GBM can be used to identify patients with poor survival trends after accounting for known confounders of GBM patient outcome.