Objective <p>The accuracy of determining gait independence using adductor pollicis muscle thickness (APMT) and skeletal muscle mass index (SMI) in community-dwelling older adults undergoing outpatient rehabilitation remains unclear. The purpose of this study was to compare the accuracy of determining gait independence using APMT and SMI in community-dwelling older adults undergoing outpatient rehabilitation.</p> Methods <p>This study included 98 older adults (mean age: 85.2&#xa0;years). Participants received outpatient rehabilitation one to three times a week. The main outcomes were gait independence (functional independence measure gait score: 6 or 7), skeletal muscle mass index (SMI), and APMT. Receiver operating characteristic (ROC) curves of APMT and SMI for gait independence were created, and cut-off values were calculated using the Youden index. Additionally, the area under the curve (AUC) s of the APMT model and the SMI model were compared using the Delong test.</p> Results <p>Among the 98 participants (male; 19, female; 79), 53 (54.1%) were in the gait independent group. The cut-off value of APMT calculated from the ROC curve was 13&#xa0;mm; the sensitivity and specificity were 67.9% and 86.7%, respectively; and the AUC was 0.800. The cut-off value of SMI calculated from the ROC curve was 4.6&#xa0;kg/m²; the sensitivity and specificity were 90.6% and 26.7%, respectively; and the AUC was 0.582. The AUC for the APMT model was significantly higher than the SMI model (<i>P</i> &lt; 0.001).</p> Conclusions <p>The results of this study show that the cut-off value of APMT for determining the gait independence was 13&#xa0;mm. In addition, APMT had a higher accuracy of determining gait independence than SMI. This indicates that measuring APMT is more useful for predicting gait independence than SMI in community-dwelling older adults undergoing outpatient rehabilitation.</p>