Background: The prevalence of diabetes has been increasing, imposing massive costs on nations. Diet and physical activity are recommended for diabetes management. Evidence suggests theory-based interventions are more efficacious than non-theory approaches. This study aimed to test the effectiveness of an integrated theoretical model-based intervention to encourage compliance for low-fat food consumption, carbohydrate counting, and physical activity in adults with type 2 diabetes. Methods: A 4-week parallel randomized control trial was conducted in Iran. Data were collected using a self-report questionnaire at baseline and 8-weeks post-intervention. This survey assessed the Theory of Planned Behaviour (TPB) constructs of attitude, subjective norm (others’ approval), and perceived behavioural control (PBC). We also assessed risk perceptions (motivational) and planning (volitional) from the Health Action Process Approach (HAPA). Furthermore, weight, body mass index, triglyceride (TG), and LDL-cholesterol were measured. Results: For both low-fat food consumption and physical activity, only planning revealed a significant improvement over time for intervention rather than control participants (F=8.78, p≤0.001 for low-fat vs. F=11.26, p≤0.001 for physical activity). For carbohydrate counting, significant effects were found for behavior (F=4.37, p=0.03), intention (F=8.14, p≤0.001), PBC (F=7.52, p≤0.001), and planning (F=4.54, p=0.03), reflecting improvements over time in the intervention participants compared to controls. Further, the effects of the intervention on behavior were partially mediated via participants’ degree of planning (B=0.10, SE=0.06, CI=0.01 to 0.26). The serum TG level was significantly reduced from pre to post-intervention for intervention rather than for control participants (F=18.69, p≤0.001). Conclusions: This intervention was promising for carbohydrate counting but did not show improvements for low-fat diet nor physical activity. Given the improvement in psychological measures and self-reported behaviour for carbohydrate counting, coupled with the findings for TG, future research is needed to demonstrate longer-term improvements. Current theory-based educational strategies can be adopted for the management of carbohydrate intake in diabetes.