Background: Early phase clinical trials are critical to enhancing therapies for children with advanced cancer. However, trial enrollment may intensify end-of-life (EOL) care. We evaluated patterns of EOL care for patients at a large cancer center. Methods: Single-center, retrospective cohort study of pediatric oncology patients, ages 6 months-21 years, who died in 2010-2014. We queried electronic medical records to assess frequencies of medical procedures (e.g. intubations), clinic visits, and hospital admissions in the last 6 months of life. We assessed timing of pediatric palliative care (PPC) consultation, initial advance care planning (ACP) discussion, and entry of do-not-attempt resuscitation (DNAR) orders in the chart, in relation to date of death. Patients enrolled in early phase clinical trials for at least 1 cycle (EP) were compared with those not enrolled (NEP), using Wilcoxon rank sum and Fisher exact tests. Results: For N = 125 patients, median age at death was 11.6 years (IQR 5.8-16.3); 46% were female; 70% were White, non-Hispanic. 26% were trial enrollees. Diagnoses included 42% solid tumors, 41% brain tumors, and 18% hematologic malignancies. Most patients had PPC consultation (83%), ACP discussions (91%), and DNAR orders (86%). EP and NEP cohorts did not significantly differ in baseline demographic or clinical characteristics, frequencies of medical procedures, or hospital admissions. EP patients had a higher median number of clinic visits than NEP patients (18.5 [16.3-27.2] vs. 14.1 [6.5-20.7], p = 0.0003) and received PPC consultation significantly closer to death than NEP patients (median days before death = 58 [16-84] vs. 85 [32-173], p = 0.04). There was no difference between EP and NEP patients in timing of initial ACP discussion or of DNAR order entry. Conclusions: Near the EOL, EP patients had more frequent clinic visits and later PPC consultation, but trial enrollment did not appear to delay ACP discussions or increase hospital resource use. These results suggest that early phase clinical trial enrollment does not substantially alter EOL care patterns for children with advanced cancer.