The　differential　evolution　algorithm　(DE)　for　pickand-placing　problem　of　multi-head　surface　mounting　machines　is　studied,　and　the　research　work　has　two　contributions　to　the　multi-head　case,　one　is　that　the　algorithm　has　a　special　mutation　operation;　the　other　is　the　crossover　operation.　Experimental　results　show　that　the　proposed　approach　gives　good　robot　assembly　time　and　less　movement　of　the　feeder　carrier　and　PCB　table.　Compared　with　genetic　algorithm　(GA),　the　new　method　requires　fewer　control　variables,　easier　to　use　and　less　time　to　receive　result,　and　the　CPU　time　is　saved　by　88%.　©　2009　IEEE.