Aim: Digital health interventions and economic incentives have shown promise in facilitating diabetes self‐management, though evidence is limited. Therefore, this study aimed to evaluate the effectiveness and cost‐effectiveness of a comprehensive app‐based diabetes self‐management programme with rewards for healthy behaviours and health outcomes. Materials and Methods: The TRIal to slow the Progression Of Diabetes (TRIPOD) study was an open‐label, parallel‐group, randomised controlled trial conducted at Duke‐NUS Medical School, Singapore. Adults with Type 2 Diabetes (diabetes), HbA1c of 7.5%‐11.0% (inclusive) and taking at least one oral diabetes medication were eligible. In total, 269 participants were randomised across three arms [Usual care (UC): 117, diabetes management programme (DMP) (intervention without rewards): 36, DMP+ (intervention with rewards): 116]. Data were analysed using intention‐to‐treat analysis with change in HbA1c at month 12 between DMP+ and UC as the primary outcome. Cost‐effectiveness of DMP+ relative to UC was also calculated. Results: Mean HbA1c improved by 0.1% in UC and by 0.5% in DMP+ at 12 months, revealing a mean difference of 0.4% (95% confidence interval (CI): −0.70, −0.08, p = 0.015). The odds ratio of HbA1c improvements of >0.5% was 2.12 (95% CI: 1.17, 3.85, p = 0.013) for DMP+ relative to UC. The incremental cost‐effectiveness ratio of DMP+ relative to UC was SGD8,516 (USD6,531) per quality‐adjusted life year gained if effectiveness could be maintained with a single year of intervention. Conclusions: A comprehensive app‐based diabetes self‐management programme with rewards for healthy behaviours and health outcomes (DMP+) cost‐effectively improved glycaemic control in Type 2 diabetes patients. Organizations focusing on value‐based healthcare should consider subsidising similar interventions.