For　the　optimization　of　multi-objective　discrete　variable　structure　optimization　in　civil　engineering,　a　new　improved　multi-objective　ant　colony　optimization　algorithm　for　truss　structures　has　been　proposed　in　this　paper.　The　new　improved　algorithm　established　two　sets,　including　sets　of　feasible　solutions　and　non-feasible　solutions,　and　＂the　repeated　solutions＂　was　replaced　with　＂the　special　solutions＂　to　acquire　the　Pareto　optimal　front-end　of　the　multi-objective　problems.　This　algorithm　can　not　only　make　the　multi-objective　ant　colony　algorithm　better　solve　multi-objective　problems　under　discrete　variables,　but　also　realize　the　successful　application　of　multi-objective　ant　colony　algorithm　in　truss　structure.　The　present　paper　proposes　a　new　improved　multi-objective　ant　colony　optimization　algorithm　for　truss　structures,　which　can　provide　a　good　performance　in　program　design,　arithmetic　speed　and　generality　of　the　proposed　method.　It　is　also　a　simple　and　practical,　and　suitable　for　projects　in　the　future.　©　Published　under　licence　by　IOP　Publishing　Ltd.