At　the　design　stage　of　a　machine　tool,　how　to　accurately　predict　and　analyze　the　motion　error　of　linear　axis　system　is　a　difficult　problem　for　designers.　This　research　is　of　great　significance　to　optimize　the　initial　design　scheme　of　the　machine　tool　and　improve　the　geometric　accuracy.　A　kinematic　errors　prediction　and　accuracy　improvement　method　of　guide　rail　based　on　assembly　and　manufacturing　analysis　was　proposed.　Firstly,　the　distribution　law　of　bolt　preload　under　a　specific　assembly　sequence　was　revealed　by　finite　element　analysis　(FEA).　Based　on　a　probability　analysis　of　assembly　parameters,　the　statically　indeterminate　mechanical　model　of　guide　rail　assembly　deformation　was　established.　Secondly,　a　Fourier　function　with　tolerance　as　amplitude　was　established　to　characterize　the　surface　topography　of　the　guide　rail　installation　base　surface　satisfying　Dirichlet　boundary　conditions.　Based　on　the　influence　of　manufacturing　and　assembly,　the　final　profile　curve　function　of　the　guide　rail　is　obtained,　and　the　straightness　and　angular　kinematics　error　(SAEs)　are　accurately　predicted.　The　prediction　accuracy　of　the　straightness　error　is　98.5%.　Referring　to　the　prediction　results,　the　targeted　accuracy　improvement　measures　for　the　linear　axis　platform　are　proposed.　The　straightness　kinematics　error　of　the　guide　rail　is　reduced　from　9.11um　to　4.28um.　The　proposed　method　can　accurately　predict　the　motion　error　of　the　guide　rail　and　guide　the　precision　improvement　of　the　linear　axis.　©　2023　SPIE.