To evaluate the effects of mirror therapy, using motor imagery training, on lower-extremity motor recovery and motor functioning of patients with subacute stroke.

Randomized, controlled, assessor-blinded, 4-week trial, with follow-up at 6 months.

Rehabilitation education and research hospital.

A total of 40 inpatients with stroke (mean age, 63.5y), all within 12 months poststroke and without volitional ankle dorsiflexion.

Thirty minutes per day of the mirror therapy program, consisting of nonparetic ankle dorsiflexion movements or sham therapy, in addition to a conventional stroke rehabilitation program, 5 days a week, 2 to 5 hours a day, for 4 weeks.

The Brunnstrom stages of motor recovery, spasticity assessed by the Modified Ashworth Scale (MAS), walking ability (Functional Ambulation Categories [FAC]), and motor functioning (motor items of the FIM instrument).

The mean change score and 95% confidence interval (CI) of the Brunnstrom stages (mean, 1.7; 95% CI, 1.2–2.1; vs mean, 0.8; 95% CI, 0.5–1.2; P=.002), as well as the FIM motor score (mean, 21.4; 95% CI, 18.2–24.7; vs mean, 12.5; 95% CI, 9.6–14.8; P=.001) showed significantly more improvement at follow-up in the mirror group compared with the control group. Neither MAS (mean, 0.8; 95% CI, 0.4–1.2; vs mean, 0.3; 95% CI, 0.1–0.7; P=.102) nor FAC (mean, 1.7; 95% CI, 1.2–2.1; vs mean, 1.5; 95% CI, 1.1–1.9; P=.610) showed a significant difference between the groups.

Mirror therapy combined with a conventional stroke rehabilitation program enhances lower-extremity motor recovery and motor functioning in subacute stroke patients.