Currently,　research　on　the　optimization　of　key　process　parameters　in　the　polishing　process　usually　uses　orthogonal　test　method,　which　needs　a　lot　of　experiments.　In　order　to　solve　this　problem,　this　paper　proposes　a　parameters　optimization　of　magnetic　composite　fluid　polishing　process　based　on　response　surface　method.　Firstly,　roughness　of　the　polycarbonate　microfluidic　chip　surface　polished　with　magnetic　composite　fluid　was　used　as　the　optimization　object,　while　the　machining　gap,　spindle　speed　and　machining　time　were　defined　as　the　key　parameters,　and　experiments　were　designed　and　conducted　based　on　response　surface　method.　Subsequently,　using　the　above　experimental　data,　a　prediction　model　for　microfluidic　chip　surface　roughness　after　polishing　that　depends　on　key　parameters　is　established.　After　that,　by　the　application　of　the　prediction　model,　the　key　parameters　were　optimized　and　verified　by　experiments.　The　results　indicated　that　by　using　the　optimized　process　parameters,　the　minimum　surface　roughness　of　the　microfluidic　chip　reduced　from　0.5　mu　m　to　0.0384　mu　m,　while　the　maximum　and　minimum　errors　between　measured　and　predicted　results　are　0.0052　mu　m　and　0.0007　mu　m,　respectively.　The　above　result　has　proved　the　accuracy　of　the　established　model　and　the　effectiveness　of　the　proposed　parameters　optimization　method.