Most　pattern　recognition　methods　used　in　condition　diagnosis　of　rotating　machinery　are　studied　that　the　sufficient　samples　are　available.　However,　it　is　hard　to　obtain　sufficient　fault　samples　in　practice.　Support　vector　machine　(SVM)　can　solve　the　learning　problem　with　a　small　number　of　samples.　This　paper　presents　a　condition　diagnosis　method　for　a　centrifugal　blower　using　a　multi-class　classification　technique,　such　as　SVM　to　identify　fault　types.　The　statistic　feature　parameters　are　also　acquired　in　the　frequency　domain　for　classification　purposes,　and　those　parameters　can　reflect　the　characteristics　of　vibration　signals.　The　effectiveness　of　the　method　is　verified　by　the　application　to　the　condition　diagnosis　for　a　centrifugal　blower.　The　result　shows　that　the　multi-class　SVM　produces　promising　results　and　has　the　potential　for　use　in　fault　diagnosis　of　rotating　machinery.