A multiple linear regression model was developed for the prediction of the 2, 7, and 28-day compressive strength of CEM I 42.5 produced in Turkey Cement Fabrics. Attention has been given to the right choice of independent variables involved in this model, especially the characteristics of the cement itself. Different combinations of variables were introduced into the model, in order to choose the variables that can properly predict the compressive strength of the cement. Multiple Linear Regression (MLR) analyses with backward stepwise were performed to describe the relationships between the compressive strength values and the chemical or physical properties of the cement. The advantage of using this technique lies in the fact that it deals simultaneously with several variables. The analysis is designed to see which factors are significant in explaining the compressive strength of the cement. The evaluation of the proposed model was performed by various statistical tests, all of which were successful. These statistical tests included: multiple correlation, test of the significance of coefficients (t-test), estimation of confidence intervals for coefficients, conditional sums of squares, R-squared and analysis of variance. Models obtained this way can predict the compressive strength of the cement with very small standard errors and coefficients of correlation of 0.9961 and 0.9955, and 0.9983, for cement strengths at 2, 7 and 28 days respectively. There was very good agreement between the strength predicted by the multiple regression model and the experimental results. These models explained only 99% of the variability in strengths.