Introduction: One of the hallmarks of cancer is reprogramming of energy metabolism whereby cancer cells become dependent on aerobic glycoysis, making high glucose uptake essential. To optimally utilize novel therapeutics targeting these processes, it is important to identify patients with tumors highly dependent on glucose uptake and to be informed about intra-patient heterogeneity. 2-deoxy-2-[fluorine-18]fluoro-D-glucose emission tomography ( F-FDG-PET) non-invasively visualizes whole body glucose uptake and could be used for this purpose. We aimed to identify a subset of non-small cell lung cancer (NSCLC) patients with high tumor glucose uptake using F-FDG-PET scans and to determine intra-patient heterogeneity and patient and tumor characteristics in this group. Methods: Primary diagnostic F-FDG-PET/CT scans of NSCLC patients were retrospectively selected from 2418 consecutive scans performed at the University Medical Center Groningen (UMCG) in 2013. Exclusion criteria were: another active malignancy and inadequate fasting prior to the scan. Included were 32 patients with concurrent type 2 diabetes mellitus (T2DM) and 32 consecutive eligible control patients. All F-FDG-PET/CT scans were performed and analyzed according to the European Association of Nuclear Medicine Research Ltd (EARL) accreditation standards ensuring standardized data collection. A region of interest (ROI) was drawn around each visible F-FDG avid tumor lesion, and in normal tissue: liver, heart, peripheral muscle and abdominal fat. Subsequently, for all ROIs, standardized uptake values (SUVs) mean 40% were determined. High tumor uptake was defined as a SUVmean >5 and very high tumor uptake as a SUVmean >8. The metabolic tumor burden was evaluated by determining whole body metabolic tumor volume (MTV), defined as the volume corresponding to regions in which the SUVmean was measured. The analysis of glucose uptake was only performed for lesions with a volume≥1 mL due to underestimation of SUV values in smaller lesions due to partial volume effects. IBM SPSS Statistics 22 was used for conducting the Chi-square tests. Results: 673 F-FDG-PET avid lesions were analyzed in 64 patients. The mean age was 63 years, 63% was male, 56% had adeno- and 39% had squamous cell carcinoma histology. Median number of tumor lesions per patient was 3 (1-90). There was a large heterogeneity in tumor glucose uptake both within and between patients. The highest measured SUVmean was 15.8 and the lowest 0.5. In the patient with the highest heterogeneity the absolute difference between the highest and lowest measured SUVmean in tumor lesions was 14. In 8/64 patients all lesions were very highly active (SUVmean >8), in 14/64 patients more than half of the lesions were very highly active and 28/64 patients had at least one very highly active lesion. In 17/64 patients all lesions were highly active (SUVmean >5) and in 28/64 patients more than half of the lesions were highly active, this was more often the case in patients with T2DM (14/32 vs 3/32 p=0.002 and 19/32 vs 9/32 p=0.012 respectively) and more often the case in patients with squamous cell carcinoma (11/25 vs 4/36 p=0.003 and 15/25 vs 10/36 p=0.012). In 46/64 patients there was at least one highly active lesion. In 18/64 patients all lesions had a SUVmean≤5. The volume of each tumor lesion (lesion MTV) and whole body tumor volume (MTV) did not correlate with tumor glucose uptake as measured using F-FDG. Conclusion: In a significant subset of NSCLC patients the majority of lesions are highly metabolically active. These patients may be especially sensitive to drugs targeting glucose uptake or glycolysis.