AIMS: This study aimed to evaluate the effectiveness of continuous glucose monitoring (CGM) with low-glucose alerts for preventing hypoglycemia in insulin-treated drivers with diabetes. METHODS: This single-center, open-label, randomized crossover study involved 30 insulin-treated participants with diabetes who drove cars at least thrice weekly in Japan. Participants underwent two 4-week periods: an alert period using CGM with active low-glucose alerts and a no-alert period using blinded CGM without low-glucose alerts, separated by an eight-week washout period. The primary outcome was the percentage of time below range (TBR
<
3.9 mmol/L). RESULTS: Twenty-seven of the 30 participants completed the CGM analysis. Although the TBR did not differ between the alert and no-alert periods among all participants, it significantly decreased during the alert period compared with the no-alert period among the participants with type 1 diabetes (-4.4 [95 % confidence interval - 8.7, -0.08]%, p = 0.047). The incidence of low-glucose when driving was significantly lower during the alert period than during the no-alert period (19 % vs. 33 %, p = 0.041). CONCLUSION: Low-glucose alerts improved the TBR in drivers with type 1 diabetes and reduced the incidence of low-glucose while driving among all insulin-treated drivers, suggesting that these alerts may ensure the safety of insulin-treated drivers.