Background: Confocal laser endomicroscopy (CLE) has advantages in detecting gastric neoplastic lesions, meanwhile it requires strict patient cooperation. Sedation could improve patient cooperation and quality of endoscopy. However, sedation is still not very popular in some resource-limited countries and regions. The purpose of this study was to compare propofol-based sedated versus un-sedated CLE in the value of diagnosing early gastric cancer (EGC) and precancerous lesions.Methods: A retrospective, cohort, single center study of 226 patients who underwent CLE between January 1, 2015 and December 31, 2017 was performed.Patients enrolled were allocated into propofol-based sedated group (n=126) and un-sedated group (n=100). The comparison of validity and reliability of CLE for identifying EGC and precancerous lesions between the two groups was performed through analyzing CLE diagnose and pathological diagnose. Results: The area under receiver operating characteristic curve (AUROC) of diagnosing EGC in sedated group was 0.97 (95% CI: 0.95 to 0.99), which was higher than that in un-sedated group (0.88 (95% CI: 0.80 to 0.97), P =0.0407). CLE with sedation performed better than without sedation in diagnosing intraepithelial neoplasia and intestinal metaplasia (P =0.0008 and P =0.0084, respectively). For patients considered as high-grade intraepithelial neoplasia or EGC by endoscopists, they would not get biopsy during CLE but receive ESD subsequently, and the misdiagnosis rate of CLE was 0 in sedated group and 27.59% in un-sedated group (P =0.006).Conclusion: Propofol based sedation was associated with improved diagnostic value of CLE for detecting EGC as well as precancerous lesions.