e18639 Background: Clinical oncology pathways (COP) provide treatment (Rx) decision support and allow detailed quality oversight. There are limited data on use of COP and the proportion of patients receiving COP recommended Rx (concordant / “On Pathway” Rx – OnP) and on the reasons for COP non-concordant “Off Pathway” Rx (OffP). Methods: COP concordance and reasons for OffP Rx were evaluated and compared at an NCI-designated cancer center (Roswell Park Comprehensive Cancer Center – RP) and its 5 affiliated community practice sites (Roswell Park Care Network – RPCN) for COP Rx decisions from 10/1/19 – 9/30/20. Elsevier ClinicalPath Oncology COPs (formerly Via Oncology) were implemented in 2018 at RP and RPCN and are completed at initiation of any systemic Rx. COPs cover about 85% of cancers, excluding most hematologic cancers. COP Rx decisions are classified as OnP (including Rx on a clinical trial) or OffP. All OffP decisions are reviewed retrospectively to identify the reasons for OffP Rx. Differences in OnP rates at RP and RPCN were compared by chi square test or Fisher’s test overall and for 7 cancer type groups and for adjuvant Rx (ADJ) and metastatic disease (MET). ADJ/MET comparison excluded gynecologic cancer (GYN), myeloma and myelodysplastic syndrome (MDS) because COPs do not classify ADJ vs MET for these. Results: At RP there were 4,291 Rx decisions for 3,361 patients and at RPCN 1,349 Rx decisions for 1,198 patients. Breast cancer was the most common cancer type (18% of patients at RP and 36% at RPCN). Proportions of patients with GYN, melanoma, sarcoma, and GU cancers were higher at RP and similar for GI, thoracic and myeloma/MDS. Among 3,153 ADJ and MET Rx decisions at RP, 55% were for METS vs. 36% of 1,096 decisions in the RPCN. Pathway decision concordance (OnP) rates overall were 85% at RP and 89% at RPCN (p < 0.01- Table). OnP rates for ADJ Rx were 89% at RP and 92% at RPCN (p = 0.05), and for MET 84% at RP and 86% at RPCN (p = 0.53). Most OffP Rx was for medically valid reasons including comorbidity, new/targeted therapies, and prior Rx precluding COP recommended Rx. Conclusions: COP concordance was slightly higher in community practice, potentially because newer therapies, patients with greater morbidities and more complex cancers are in the academic practice. RP had a higher proportion of MET cases consistent with the tertiary practice treating more advanced cancer. OffP Rx was higher with MET, likely related to the nuances of Rx of MET cancer and new therapies not yet in pathways. OffP Rx was fully justified in most cases.[Table: see text]