The problems of small sample,non-linearity,high dimensions and local minimal value can be well solved by supporting vector machine in soft-sensor modeling.In consideration of the online measurement of the component content in rare earth countercurrent extraction separation process,algorithm of SVM with RBF kennel is applied to the modeling of the rare earth extraction separation process.The selection and effect of model-parameters are discussed.Through the simulations of the model,it shows that the component content soft-sensor model based on SVM has both preferable generalization and high velocity.SVM is an effective method for rare earth extractingprocess soft-sensor.