This research discussed, the process of comparison between the regression model of partial least squares and tree regression, where these models included two types of statistical methods represented by the first type "parameter statistics" of the partial least squares, which is adopted when the number of variables is greater than the number of observations and also when the number of observations larger than the number of variables, the second type is the "nonparametric statistic" represented by tree regression, which is the division of data in a hierarchical way. The regression models for the two models were estimated, and then the comparison between them, where the comparison between these methods was according to a Mean Squares Error (MSE) and using the simulation of the experiment and by taking different sample sizes. where the results of the simulation showed that the regression of partial least squares is best when taking the following contrast variance values (0.01, 0.5, 1) and for all sample sizes, whereas tree regression is the best when it is The variance value is large (5) and for all sample sizes.
