An　automatic　CCPs　(critical　control　points)　identification　method　for　lettuce＇s　fresh-cut　processing　HACCP　(hazard　analysis　and　critical　control　point)　implementation　based　on　optimized　SVM　(support　vector　machine)　model　was　introduced.　The　CCPs　identification　tests　which　use　the　general　SVM　algorithm　showed　its　weakness　because　of　the　man-made　kernel　selection　and　the　classification　results　had　larger　fluctuation.　Generic　algorithm　was　used　to　obtain　an　optimized　kernel　function　and　then　the　GA-SVM　model　was　proposed.　The　model　had　been　applied　in　the　CCPs　identification　of　lettuce＇s　fresh-cut　processing　and　the　classification　accuracy　of　87.5%　was　stabler　than　the　traditional　method.　The　method　is　also　general　which　can　be　easily　popularized　to　other　HACCP　implementations.