Funding Acknowledgements Type of funding sources: None. Introduction Diagnosis of blood culture negative infective endocarditis (BCNIE) is challenging, with positive microbiology key in the modified Duke Criteria (mDC). ESC IE Guidelines (2015) recommend the use of 18F-FDG PET/CT (PET) in cases where transoesophageal echocardiography (TOE) is equivocal and where cardiac implantable device-related IE (CIDRE) is suspected. We explored the role of PET to improve diagnostic certainty in the challenging cohort of BCNIE. Methods Retrospective review of all suspected BCNIE patients undergoing PET (10/2015 to 01/2021). Myocardial suppression technique was used in all cases, and studies were assessed for valve/device avidity. Patients were classified as definite/possible/rejected IE by mDC pre- and post-PET, with incremental benefit assessed by net reclassification index (NRI) versus actual diagnosis. Actual diagnosis was defined by Endocarditis Team consensus or surgical specimen (where available) at a minimum of 2-months following index admission. Results PET was performed in 110/807 (13.6%) cases overall. BCNIE prevalence was 18% across the total cohort, with 25/110 (22.7%) PET studies in BCNIE patients (male = 17, mean age 65). (p = 0.16 for PET in BCNIE compared to overall). PET was undertaken in 8 CIDRE, 9 native IE (NVE) and 10 prosthetic IE (PVE); 2-patients had suspected CIDRE + PVE. TOE was performed in all cases, and surgery was required in 8/25 patients. IE was confirmed in 44% of cases. PET sensitivity, specificity, positive and negative predictive values were 73%, 93%, 89% and 81%, respectively. Addition of PET to the mDC improved re-classification to definite or rejected IE, with NRI 0.633 (positive NRI 0.3; negative 0.333), and added weight to the original mDC classification in 14/25 (56%) of cases. Conclusion PET improves diagnostic certainty when combined with mDC in the evaluation of patients with BCNIE across NVE, PVE and CIDRE.