The application of the mathematical tools such as multiple linear regression analysis (MLR) in the prediction of mechanical behavior of cement-based materials is the subject of several previous investigations works conducted in civil engineering field. As new alternative of the development of green cement mortar, this work aims in the first to study the effect of the substitution of Ordinary Portland Cement (OPC) with ground granulated blast furnace slag (GGBFS) on the compressive and flexural strength of manufactured mortar. Both mechanical tests are caried out using cubic and prismatic specimens with dimensions of 4 cm × 4 cm × 4 cm and 4 cm × 4 cm × 16 cm, respectively. In the other hand, this study investigates the capacity of the multiple linear regression analysis in the prediction of compressive and flexural strength of green cement mortars. Indeed, tests samples of 64 are examined in order to develop an adequate mathematical model for the prediction of the both investigated mechanical properties as a function of the three independents variables of water/binder ratio, hydration ages and the concentration of ground granulated blast furnace slag. The mechanical tests revealed that 20 mass % of ground granulated blast furnace slag content give good properties for the development cement mortars and above this concentration there is a slight effect on the mechanical strength. According to the statistical results, this study concluded also that the multiple linear regression analysis has a good performance for the prediction of flexural strength properties than compressive strength of green cement mortars, 0.8 and 0.42 of coefficient of determination (R2) are obtained, respectively. In addition, the experimental results revealed also that the hydration age has more effect than the two others parameters which is in accordance with statistical results obtained from the MLR analysis, such as significance level (Sig.), standardized coefficients, t-test values and the variation inflation factor (VIF). Both regression models obtained for compressive and flexural strength are statically significant with very low values of sig. < 5 %. The statistical parameters indicate clearly that the hydration age has a strong effect on the prediction of compressive and flexural strength values using the obtained models, which mainly due to its highest values of standardized coefficients of 0.61. The results of these parameters indicate that water/binder ratios have more effect than ground granulated blast furnace slag content with standardized coefficients values of 0.13 and -0.17, respectively, but this is not the same for the flexural strength values, the concentration of ground granulated blast furnace slag has more effect than water/binder ratios, -0.22 and -0.66, respectively.