This　paper　proposes　a　heuristic　multi-phase　approach　to　minimize　the　moving　distance　of　gantries　by　balancing　the　workload　of　a　dual-delivery　surface　mount　technology　(SMT)　placement　machine.　The　SMT　placement　machine　under　study　is　a　rotary-head　dual-gantry　machine.　Several　practical　factors　are　considered,　such　as　the　component-nozzle　compatibility　and　the　machine　structure.　The　machine　has　two　stations　with　a　symmetric　layout.　Each　station　has　a　printed　circuit　board　(PCB)　conveyor　belt,　a　fixed　camera,　an　auto　nozzle　changer,　a　feeder　base,　and　a　movable　gantry　with　several　nozzle　heads.　In　the　pick-and-place　assembly　operation,　two　independent　gantries　alternately　mount　on　one　PCB.　Most　research　in　literature　considers　three　main　decisions　in　the　SMT　placement　machine　optimization:　nozzle　setup,　feeder　arrangement,　and　pick-and-place　sequence.　In　this　research,　two　more　decisions　are　introduced　into　the　dual-gantry　problem:　workload　balance　between　gantries　and　gantry　cycle　scheduling.　A　hierarchical　strategy　is　developed　to　solve　the　workload　balance　problem,　including　nozzle　and　component　allocations.　The　other　decisions,　such　as　the　feeder　arrangement　and　the　pick-and-place　sequence,　are　made　using　existing　heuristics.　The　experimental　results　show　that　this　heuristic　approach　has　advantages　compared　to　other　algorithms　proposed　in　literature.