Abstract Extrachromosomal circular DNA (ecDNA) is an important driver of aggressive cancers, including medulloblastoma (MB), the most common malignant pediatric brain tumor. To assess the clinical import of ecDNA in MB, we applied computational methods to detect ecDNA to the genomes of a cohort of 468 MB patients and 31 MB model systems. Among patients, ecDNA was detected in 18% of tumors and carried threefold greater risk of mortality. Between sequential patient biopsies at initial diagnosis and subsequent relapse, we observed frequent structural variation at ecDNA loci and generation of new ecDNA sequences. Among model systems, ecDNA was found in 19 of 31 genomes (61%). Although ecDNA was far more prevalent among MB models than patients, the ecDNA genomic sequence was conserved between most patient-derived xenograft (PDX) models and their origin human tumors. EcDNA promotes high oncogene copy number, accelerated evolution of drug resistance and poor prognosis. Mechanistically, these behaviors are hypothesized to be consequences of intratumoral heterogeneity of ecDNA copy number; however, there are currently no methods to quantify the abundance and distribution of ecDNA across single cells in a tumor. We have previously identified a primary TP53-mutant SHH medulloblastoma tumor, which harbors multiple distinct lineages of ecDNA originating from different chromosomes. We have now simultaneously analyzed the accessible chromatin and gene expression in single cells of the human tumor and of a orthotopic mouse xenograft (PDX) derived for this tumor, using multiome single-cell ATAC-seq and gene expression (10X Genomics). By measuring scATAC-seq read depth at the ecDNA loci relative to the rest of the genome, we estimate ecDNA copy number in single cells. We observe substantial variation in the ecDNA copy number of cells in both the human tumor and the PDX, and strong association with transcription of ecDNA-amplified genes. We anticipate that this approach will illuminate mechanisms of ecDNA-mediated drug resistance, by revealing the sensitivity and population dynamics of ecDNA-containing cells under treatment pressure. Citation Format: Owen Chapman, Sunita Sridhar, Jon D. Larson, Robert J. Wechsler-Reya, Jill Mesirov, Lukas Chavez. Detecting and characterizing extrachromosomal DNA in medulloblastoma at single-cell resolution [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 6214.