Brainstem toxicity after radiation is a devastating complication and a particular concern with proton radiation (RT), given the biological differences and range uncertainties at the distal edge of proton beam. Data from St. Jude's found a 3.7% risk of brainstem necrosis after photon-based RT, while the incidence with protons ranges from 0.7% - 31% across studies. We investigated the incidence and clinical correlates of brainstem injury in pediatric brain tumors treated with proton therapy.All patients < 18 years with infratentorial brain and pineal tumors, treated with proton beam RT at our institution from 2007 - 2019, with a brainstem Dmean of > 30 Gy and/or Dmax > 50.4 Gy were included. We excluded patients with primary brainstem tumors. Symptomatic brainstem injury (SBI) was defined as any new or progressive cranial neuropathy, ataxia, dysmetria and/or motor weakness with corresponding radiographic abnormality within the brainstem and was graded per the NCI Common Terminology Criteria for Adverse Events v5.0 (Grade 2: moderate symptoms, corticosteroids indicated; Grade 3: severe symptoms, medical intervention indicated; Grade 4: life-threatening; urgent intervention; and Grade 5: death).A total of 595 patients were reviewed and 362 (medulloblastoma = 209, ependymoma = 86, ATRT = 43, pineoblastoma = 9, others = 15) met our inclusion criteria. Median age at RT was 5 years (range 0.7 - 17.9 years) and median prescribed RT dose was 54 CGyE (range 39.6 - 59.4 CGyE). Median follow up was 40 months (range 1 - 152). Ten patients (2.7%) developed SBI, at a median of 4 months after RT (range 1-6). Grades 2,3,4 and 5 brainstem injuries were seen in 3, 4, 2 and 1 patient respectively. Mean brainstem Dmax in SBI and non-SBI patients was 55.9 Gy and 54.8 Gy (P = 0.30), while the mean Dmean was 43.3 Gy and 43.2 Gy (P = 0.48) respectively. Most common symptoms of SBI were ataxia, slurred speech, 6th cranial nerve palsy and bulbar palsy. All SBI patients required steroids for 1-5 months, while grade 3-5 injuries were also treated with bevacizumab and/or hyperbaric oxygen therapy. Asymptomatic imaging changes within brainstem were seen in 47 patients (13%) at a median of 4 months after RT, most common findings being increased T2 flair, focal enhancement and encephalomalacia. In 2014, our institution started using strict brainstem dose constraints (Dmax ≤ 57 Gy, Dmean ≤ 52.4 Gy and V54 ≤ 10%). The incidence of SBI has decreased from 4.2% (2007-2013) to 0.7% (2014-2019) (P = 0.047) while asymptomatic imaging changes remain similar (11.6% vs 15.1%, P = 0.33). Patients with SBI had a higher V45, V50-52 and D15 -D50; while the Dmax, Dmean, V53-60 and D1- D10 were similar.Our results suggest that there is a low risk of SBI after proton therapy for pediatric brain tumors. A lower incidence of SBI was seen after the use of strict brainstem dose constraints. Volumetric doses to brainstem may play a significant role in determining the risk of SBI and developing a dose response nomogram is warranted.