Abstarct: The main aim of this work was to development of a quantitative structure-property relationship(QSPR) model for isobaric expansivity prediction of organic fluids.The large number of descxriptors were calculated with Dragon software and the best calculated descxriptors were selected from them with data splitting-feature selection (CDFS) strategy. In the CDFS strategy, data splitting is achieved many times and in each case feature selection is performed using stepwise regression (SR) method. The resulted models are compared for similarity and dissimilarity between the selected descxriptors. The final model is one whose descxriptors are the common variables between all of resulted models. The 6 descxriptors with CDFS strategy were selected. The selected descxriptors with the feature selection method and two experimental variables (T and p) were used as input for constructing multiple linear regression model, artificial neural network and support vector machines. The performance of each model was investigated by test set. The results showed that the support vector machine model, may be simulated the relationship between the structural descxriptors and the isobaric expansivity of the desired molecules accurately. The mean square errors (MSEs) for the test set are 8.4×10-5. Also, the performance of the nonlinear models was compared with multi linear regression (MLR) model. The results indicates that superiority of the nonlinear models over that of the linear MLR model.