Diffuse intrinsic pontine glioma (DIPG) is a highly aggressive childhood tumor that is nearly uniformly fatal. Treatment remains challenging, and criteria for response assessment are poorly implemented in DIPG due to inter-observer variability in tumor volume measurement, non-enhancing components on MRI, and undefined relationships between radiographic parameters and survival. We longitudinally evaluated pontine and brainstem volumes in children with DIPG to test the hypothesis that 3D measurements of these structures can serve as reproducible biomarkers of local disease burden and response to therapy. Thirty-two consecutive patients with newly diagnosed DIPG were treated with standard definitive radiation therapy from 2010 – 2016 at a single institution. Using built-in knowledge-based segmentation software, we generated semi-automated 3-D measurements of the pons and brainstem from standard T1-weighted MR images and then manually adjusted volumes based on anatomical landmarks. Volumetric measurements were collected longitudinally throughout each patient’s clinical course. We calculated overall survival (OS) using the Kaplan-Meier method and compared groups using the log rank method. A p-value < 0.05 was considered significant. Median follow up was 8.0 months, and median OS from end of radiation for all patients was 12.2 months. The median age at diagnosis was 5.6 years (range 2 to 11 years). The median pontine and brainstem volumes at diagnosis were 36.0 ± 1.9 mL and 50.1 ± 2.2 mL, respectively. Radiation was associated with a greater decrease in pontine volume as compared to brainstem volume at early post-treatment time points (-41.6 ± 5.5% vs. -21.6 ± 3.9%, respectively, at one month, p = 0.004). By time of last follow up, pontine and brainstem volumes had increased to 29.9 ± 2.7 mL and 48.1 ± 3.2 mL, no longer significantly different from pre-treatment volumes (p = 0.20 and p = 0.43, respectively). There was strong agreement between paired readers for pontine volumes (median difference 6.9% with median coefficient of variation 4.9%). Patients whose pontine volume decreased by more than 40% at one-month post-radiation had a non-significant trend toward increased OS (14/27 patients; 13.8 ± 6.3 vs. 10.1 ± 0.9 months, p = 0.09). Volumetric changes in the pons and brainstem occur in response to treatment, correlate with local disease burden, and can be measured reproducibly. These volumetric analyses may serve as useful biomarkers in ongoing and future trials in children with DIPG involving localized delivery of therapeutics, e.g. convection-enhanced delivery of radioimmunotherapies, that require evaluation of locoregional disease control in addition to survival.