Objective. To assess the efficacy of medial-wedge insoles in valgus knee osteoarthritis (OA). Methods. Thirty consecutive women with valgus-deformity knee OA >8 degrees were randomized into 2 groups: medial insole (insoles with 8-mm medial elevation at the rearfoot [n ؍ 16]) and neutral insole (similar insole without elevation [n ؍ 14]). Both groups also wore ankle supports. A blinded examiner assessed pain on movement, at rest, and at night with a visual analog scale (VAS), the Lequesne index, and Western Ontario and McMaster Universities Osteoarthritis (WOMAC) Index. Femorotibial, talocalcaneal, and talar tilt angles were evaluated at baseline and after 8 weeks of insole use. Results. Significant reductions in the medial insole group were observed for pain on movement (mean ؎ SD VAS pre-and postintervention 8.1 ؎ 1.5 versus 4.2 ؎ 2.4; P ؍ 0.001), at rest (5.1 ؎ 2.3 versus 2.7 ؎ 2.4; P ؍ 0.002), and at night (6.1 ؎ 2.7 versus 3.1 ؎ 2.1; P ؍ 0.001). In addition, a decrease in Lequesne (14.7 ؎ 3.4 versus 9.6 ؎ 3.8; P ؍ 0.001) and WOMAC scores (74.1 ؎ 14.2 versus 56.1 ؎ 14.9; P ؍ 0.001) was observed for the medial insole group. In the neutral insole group, a significant reduction was observed only for night pain (mean ؎ SD VAS pre-and postintervention 5.8 ؎ 2.4 versus 4.6 ؎ 2.4; P ؍ 0.019). An increase in femorotibial angle (169.0 ؎ 3.4 versus 170.8 ؎ 3.7; P ؍ 0.001) occurred only in the medial insole group. Moreover, the difference in measured femorotibial angles pre-and postintervention was 1.84 ؎ 1.42 versus ؊0.18 ؎ 0.67 (P < 0.001) for the medial and neutral insole groups. Conclusion. The use of medial-wedge insoles was highly effective in reducing pain at rest and on movement and promoted a functional improvement of valgus knee OA.