Objective To evaluate the validity of Community Extremity Disability Assessment Schedule (CEDA) in community rehabilitation. Methods 313 patients with hemiplegia after stroke were divided into rehabilitation group and control group. The rehabilitation group received community-based rehabilitation (CBR) for 3 months, and was assessed with Community Extremity Disability Assessment Schedule before and 1 month, 2 months and 3 months after rehabilitation respectively. And it was compared with Fugl-Meyer Assessment (FMA). The control group received the same evaluation only. Results The scores of CEDA and FMA improved more in the rehabilitation group than in the control group 2 months after rehabilitation (P<0.01). The score of CEDA positively correlated with the FMA (r=0.643, P<0.01). Conclusion CEDA can be used to evaluate the efficacy of community rehabilitation.