Abstract Background Patients with cardiac sarcoidosis (CS) face an elevated risk of fatal ventricular arrhythmic events (FVAE), including sudden cardiac death. Late gadolinium enhancement (LGE) from contrast-enhanced cardiac magnetic resonance is used for diagnosis of CS and estimation of FVAE risk. However, its application is limited in cases with cardiac devices, renal failure or contrast agent allergies. 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) emerges as a diagnostic tool for CS, even in cases where LGE is not feasible. Purpose This study aimed to investigate the potential utility of FDG-PET in predicting FVAE in patients with CS by hypothesizing that higher or lower percent FDG uptake could predict the FVAE risk. Methods Cardiac FDG-PET data from 121 patients with CS, acquired at our university hospital between March 2008 and November 2020, were analyzed. A polar-map model was used to calculate percent uptake relative to maximal cardiac FDG uptake in each of the American Heart Association (AHA) 17 segments (Figure A). Percent uptakes of the AHA 17 segments were compared between groups with and without future FVAE. LGE information was available in 82 patients and was used to assess the presence of LGE in regions where percent FDG uptake differed between the two groups. Results Among enrolled patients (mean age 59.5 ± 10.0 years, 67.8% women), 43 experienced FVAE following FDG-PET image acquisition. The group with future FVAE had a higher history of FVAE (7.7% vs. 37.2%) and beta-blocker administration (38.5% vs. 60.5%), larger left ventricular diastolic diameter (LVDd) (49.3 ± 7.3 mm vs. 57.4 ± 9.5 mm) and lower ejection fraction (LVEF) (56.6 ± 13.6% vs. 42.0 ± 15.2%) compared to the group without it (P < 0.05 all). Percent FDG uptake in the basal interventricular septum part (segment 3) (Figure A) was lower in patients with future FVAE compared to those without it (54.4 ± 13.8% vs. 41.8 ± 15.2%, P < 0.001 corrected for multiple comparisons). Kaplan-Meier analysis with a median split of percent FDG uptake in segment 3 showed that patients with lower uptake (<49%, N= 62) had lower FVAE-free survival compared to those with higher uptake (≥49%, N = 59) (P = 0.007) (Figure B). A Cox hazard model also demonstrated an association between percent FDG uptake in segment 3 and FVAE onset (hazard ratio 0.972 [95% CI 0.951, 0.995], P = 0.015) after adjusting for previous FVAE, beta-blocker administration, LVDd and LVEF. Similarly, analysis within the subgroup with LGE in the septum (N = 58) revealed a higher risk of FVAE in the group with lower percent FDG uptake compared to those with higher uptake (P = 0.020) (Figure C(a)). However, this association was not evident in patients without LGE in the septum (N = 24) (P = 0.821) (Figure C(b)). Conclusions Reduced percent FDG uptake with LGE in the basal intraventricular septum may serve as a valuable predictor of future FVAE occurrence in patients with CS.