Large　scare　redundant　data　in　heterogeneous　information　integration　systems　creates　computational　challenges.　In　order　to　optimize　the　join　order,　a　multi-copy　join　query　optimization　based　on　genetic　algorithm　is　proposed.　Choosing　redundant　copies　of　tables　and　searching　for　a　good　join　order　were　represented　by　fitness　function　and　genetic　operators　as　well.　Experimental　results　show　that　the　multi-copy　join　query　optimization　is　computationally　efficient　and　necessary　in　information　integration　system.