OBJECTIVE: To evaluate the effect of implementation intentions as an intervention strategy to promote walking in adults with type 2 diabetes mellitus (T2DM). METHODS: We conducted a controlled and randomized trial, with 12 months of follow-up, involving 65 people with T2DM recruited from primary health care units and allocated them in the control group (CG, n = 32) and intervention group (IG, n = 33). The IG received the implementation intention strategy to promote walking and the CG remained in follow-up for conventional treatment in primary health care. The researchers were blinded by anthropometric measurements and the filling of the instruments. RESULTS: After twelve months of follow up, the IG presented a statistically significant increase in the leisure time physical activity when compared with the CG (p = 0.0413) and showed a significant decrease in waist circumference (p = 0.0061). No significant difference was observed regarding body mass index and glycated hemoglobin among groups. CONCLUSIONS: Implementation intention was effective in promoting walking and improving clinical indicators in adults with T2DM.