Machining　accuracy　is　critical　for　the　quality　and　performance　of　a　mechanical　product,　and　the　reliability　of　a　multi-axis　NC　machine　tool　reflects　the　ability　to　reach　and　maintain　the　required　machining　accuracy.　The　objective　of　this　study　is　to　propose　a　general　methodology　that　will　simultaneously　consider　geometric　errors　and　thermal-induced　errors　to　allocate　the　geometric　accuracy　of　components,　for　improving　machining　accuracy　reliability　under　certain　design　requirements.　The　multi-body　system　(MBS)　theory　was　applied　to　develop　a　comprehensive　volumetric　error　model,　showing　the　coupling　relationship　between　the　individual　errors　of　the　components　of　this　machine　tool　and　their　volumetric　accuracy.　Additionally,　a　thermal　error　model　was　established　based　on　the　neural　fuzzy　control　theory　and　was　compared　to　the　common　thermal　error　modeling　method　called　BP　neural　network.　Based　on　the　traditional　cost　model　and　the　reliability　analysis　model,　a　geometric　error-cost　model　and　a　geometric　error-reliability　model　were　established,　taking　the　weighted　function　principle　into　consideration.　Then,　an　allocation　approach　of　the　geometric　errors,　for　optimizing　total　cost　(manufacture　and　QLF)　and　reliability,　subject　to　the　geometrical　and　operational　constraints　of　the　machine　tool,　was　proposed　and　formulated　into　a　mathematical　model,　in　order　to　perform　the　optimization　process　of　accuracy　allocation　by　using　the　advanced　NSGA-II　algorithm.　A　case　study　was　also　performed　in　a　five-axis　machine　tool,　and　the　traditional　NSGA　algorithm　was　used　for　comparison.　The　optimization　results　for　the　five-axis　machining　center　showed　that　the　proposed　approach　is　effective　and　able　to　perform　the　optimization　of　geometric　accuracy　and　improve　the　machining　accuracy　and　the　reliability　of　the　machine　tool.