The aim of this trial was to compare the efficacy of real-time and intermittently scanned continuous glucose monitoring (rtCGM and isCGM, respectively) in maintaining optimal glycemic control.In this randomized study, adults with type 1 diabetes (T1D) and normal hypoglycemia awareness (Gold score <4) used rtCGM (Guardian Connect Mobile) or isCGM (FreeStyle Libre) during 4 days of physical activity (exercise phase) and in the subsequent 4 weeks at home (home phase). Primary end points were time in hypoglycemia (<3.9 mmol/L [<70 mg/dL]) and time in range (3.9-10.0 mmol/L [70-180 mg/dL]). The isCGM group wore an additional masked Enlite sensor (iPro2) for 6 days to check for bias between the different sensors used by the rtCGM and isCGM systems.Sixty adults with T1D (mean age 38 ± 13 years; A1C 62 ± 12 mmol/mol [7.8 ± 1.1%]) were randomized to rtCGM (n = 30) or isCGM (n = 30). All participants completed the study. Percentage of time in hypoglycemia (<3.9 mmol/L [<70 mg/dL]) was lower among rtCGM versus isCGM participants in the exercise phase (6.8 ± 5.5% vs. 11.4 ± 8.6%, respectively; P = 0.018) and during the home phase (5.3 ± 2.5% vs. 7.3 ± 4.4%, respectively; P = 0.035). Hypoglycemia differences were significant and most notable during the night. rtCGM participants spent more time in range (3.9-10 mmol/L [70-180 mg/dL]) than isCGM participants throughout both the exercise (78.5 ± 10.2% vs. 69.7 ± 16%, respectively; P = 0.0149) and home (75.6 ± 9.7% vs. 67.4 ± 17.8%, respectively; P = 0.0339) phases. The results were robust to the insignificant bias between rtCGM and isCGM sensors that masked CGM found in the isCGM arm.rtCGM was superior to isCGM in reducing hypoglycemia and improving time in range in adults with T1D with normal hypoglycemia awareness, demonstrating the value of rtCGM alarms during exercise and in daily diabetes self-management.