The　present　paper　presented　a　fast　and　non-destructive　method　for　the　discrimination　of　minnan　oolong　tea　varieties　by　near-infrared　spectroscopy　technology.　Two　hundred　ten　samples　including　Tieguanyin,　Huangjingui,　Benshan,　Maoxie　and　Meizhan　were　collected　in　different　tea　plantations　of　Minnan.　NIR　spectra　of　1　100~1　300　nm　and　1　640~2　498　nm　were　successfully　obtained.　Prediction　model　was　built　by　principal　component　analysis　(PCA),　and　the　effects　of　multiplicative　scatter　correction(MSC)　and　standard　normal　variate(SNV)　on　the　model　were　observed　and　compared.　It　was　indicated　that　the　effect　of　MSC　on　the　model　was　superior　for　the　effect　of　SNV　because　the　classification　accuracy　of　model　for　the　calibration　samples　reached　96%,　and　this　number　to　the　prediction　samples　was　about　90%.　These　results　demonstrated　that　the　near-infrared　spectroscopy　method　established　could　be　an　efficient　and　accurate　way　for　the　discrimination　of　minnan　oolong　teas　and　would　have　a　strong　practical　value.