Background: Immune checkpoint inhibitors (ICI) have resulted in promising outcomes in systemic cancer but response in glioblastoma (GBM) has not met early expectations based on early clinical trial results. Conflicting data exists regarding the incidence of pseudoprogression after ICI in brain tumors. Here, we present our institutional experience of the radiographic response following ICI therapy in GBM both at initial diagnosis and recurrence. Methods: We retrospectively identified IDH-wildtype GBM (per 2021 WHO criteria) patients treated either at time of initial diagnosis or at recurrence with nivolumab, pembrolizumab, or durvalumab between 2015 and 2022 at MGH. Overall radiographic and clinical responses after initiation of ICI were assessed in both groups. Results: 56 pts (18 female, 38 male) were identified (average age 58.6). Eight (14.3%) patients had ICI-related non-hematological toxicities, including granulomatosis folliculitis, arthritis and myositis, headache, acneiform rash, transaminitis, and gastritis). Four (7.1%) of the eight patients had to discontinue treatment due to ICI-related severe toxicity (hepatitis and/or gastritis) and were not included in the final response assessment. Of the remaining 52 patients, none had evidence of pseudoprogression related to ICI. We did not identify any complete response (CR) or partial response (PR) secondary to ICI monotherapy. In the 13 patients treated at initial diagnosis, the following radiographic responses were seen: CR=0, PR=0, stable disease (SD)=5 (38.5%), and progressive disease (PD)=8 (61.5%). In the 39 patients treated at disease recurrence, the responses were: CR=0, PR=4 (10.3%), SD=10 (33.3%), PD= 22 (56.4%). Noteworthy, all patients with PR were concurrently treated with bevacizumab. Conclusions: 0/52 patients showed radiographic response after ICI monotherapy. While treatment was overall tolerated, ICI therapy was discontinued due to toxicity in 7% of patients. Pseudoprogression attributable to ICI was not seen in any of the patients treated.