The　paper　proposes　a　framework　to　estimate　the　software　project　success　potential　using　association　rule　mining　technique.　The　framework　aims　to　explore　the　relationship　between　the　risk　dimensions　and　project　outcome.　By　the　obtained　association　rules　that　take　risk　dimensions　as　the　condition　and　the　project　outcome　as　the　result,　project　managers　can　estimate　whether　a　new　project　will　be　successful　or　failure　based　on　its　risk　factor　values　as　early　as　possible.　At　last　we　design　a　project　data　generator　to　illustrate　how　the　framework　works.　©2009　IEEE.