PURPOSE We aimed to evaluate the efficacy and safety of use of the Fasting Algorithm for Singaporeans with Type 2 Diabetes (FAST) during Ramadan. METHODS We performed a prospective, multicenter, randomized controlled trial. The inclusion criteria were age ≥21 years, baseline glycated hemoglobin (HbA 1c) level ≤9.5%, and intention to fast for ≥10 days during Ramadan. Exclusion criteria included baseline estimated glomerular filtration rate <30 mL/min, diabetesrelated hospitalization, and short-term corticosteroid therapy. Participants were randomized to intervention (use of FAST) or control (usual care without FAST) groups. Efficacy outcomes were HbA 1c level and fasting blood glucose and postprandial glucose changes, and the safety outcome was incidence of major or minor hypoglycemia during the Ramadan period. Glycemic variability and diabetes distress were also investigated. Linear mixed models were constructed to assess changes. RESULTS A total of 97 participants were randomized (intervention: n = 46, control: n = 51). The HbA 1c improvement during Ramadan was 4 times greater in the intervention group (-0.4%) than in the control group (-0.1%) (P = .049). The mean fasting blood glucose level decreased in the intervention group (-3.6 mg/ dL) and increased in the control group (+20.9 mg/dL) (P = .034). The mean postprandial glucose level showed greater improvement in the intervention group (-16.4 mg/dL) compared to the control group (-2.3 mg/dL). There were more minor hypoglycemic events based on self-monitered blood glucose readings in the control group (intervention: 4, control: 6; P = .744). Glycemic variability was not significantly different between the 2 groups (P = .284). No between-group differences in diabetes distress were observed (P = .479). CONCLUSIONS Our findings emphasize the importance of efficacious, safe, and culturally tailored epistemic tools for diabetes management.