Objective: To examine the effects of the provision of a cane, delivered to ambulatory people with chronic stroke, for improving walking and social participation. Design: Two-arm, randomized trial. Setting: Community-based. Participants: Ambulatory individuals with chronic stroke. Interventions: The experimental intervention was the provision of a single-point cane during one month. The control group received a placebo intervention. Outcome measures: Walking speed, step length, cadence, walking capacity, and walking confidence were measured without the cane to examine its rehabilitative effect. Walking speed was also measured with the cane for inclusiveness, and social participation was measured for examining carry over effects. Outcomes were measured at baseline, and after one and two months. Results: Fifty individuals were included. In the experimental group, mean age was 69 years (SD 14), and walking speed was 0.58 m/s (SD 0.17). In the control group, mean age was 68 years (SD 13), and walking speed was 0.63 m/s (SD 0.15). When walking without the cane, after one and after two months, there were no between-group differences in any measures. When walking with the cane, after one month, the experimental group walked 0.14 m/s (95% CI 0.05–0.23) faster than the control group and after two months, they were still walking 0.18 m/s (95% CI 0.06–0.30) faster. Conclusion: Use of a cane improved walking speed, only when participants walked with the cane. Use of cane for one month did not improve walking outcomes, when walking without the cane. People with stroke would need to continue to use the cane to maintain any benefits in walking speed.