A QSAR study of the antiproliferative activity is applied to a set of 22 molecules using the principal component analysis (PCA), multiple linear regression (MLR) (R = 0,754), multiple nonlinear regression (MNLR) (R = 0,981) and artificial neural network (ANN) (R = 0,987) method, and finally the model was validated with the cross-validation "leave-one-out" (CV-LOO) (R = 0,601), the values of the predicted activities are in agreement with the experimental results. From the results of this study it can be said that this model gives statistically significant results and shows a good stability to the variation of data in the cross-validation leave-one-out.