A　pattern　recognition　machine　with　fuzzy　clustering　analysis　(PRM)　is　proposed　in　this　paper,　which　is　with　a　twofold　task,　learning　knowledge　from　the　sample　patterns　and　applying　the　knowledge　to　the　observed　patterns.　With　the　same　fuzzy　clustering　procedure,　the　PRM　does　its　twofold　work.　By　fuzzy　clustering　analysis　on　the　sample　patterns　in　the　pattern　space,　the　PRM　establishes　its　clustering　criteria,　the　sample　cluster　structure　and　the　expected　cluster　structure.　The　observed　pattern　is　added　to　the　sample　pattern　set　to　form　the　extended　sample　set,　and　the　same　clustering　scheme　for　the　sample　pattern　set　is　applied　to　the　extended　sample　set　by　the　PRM　so　that　the　observed　pattern　is　recognized　under　the　clustering　criteria.　The　paper　examines　the　PRM　theoretically　and　develops　some　Significant　conclusions.　At　the　end　of　this　paper,　an　illustrative　example　is　given　to　show　how　the　PRM　works.