This　paper　discussed　a　take-off　posture　optimization　method　of　the　Locust-like　Hopping　Robot　using　particle　swarm　optimization.　The　kinematic　model　was　first　established　based　on　DH　method　for　the　locustlike　hopping　robot,　and　the　mapping　relation　of　robot　from　the　center　of　mass　space　to　the　joint　space　was　obtained.　The　jumping　performance　of　the　hopping　robot　was　evaluated　by　computing　the　velocity　and　force　directional　manipulability　of　the　hopping　robot　during　the　take-off　stage,　and　the　joint　parameters　of　the　hopping　robot　were　obtained　by　particle　swarm　optimization.　Simulation　results　show　the　optimization　method　is　feasible　and　effective.　©　2015　Technical　Committee　on　Control　Theory,　Chinese　Association　of　Automation.