One of the most important problems in concrete industry in Iraq is deterioration due to internal sulfate attack that causes damage of concrete and hence reduces its compressive strength, increases its expansion and may be lead to its cracking and destruction. Linear regression analysis is used to predict the optimum SO3 content
(O.G.C) on the basis of cement chemical composition, Blaine fineness and age. Three models are presented, the first one is an early age model (less than or equal to 7- days). Then a late age (greater than 7-days) model was developed based on the predicted optimum SO3 content of early age and late age. The third model was an all ages model and it is a general model specially for OPC. The important results
obtained are the positive effect of C3S, C3A and C4AF on optimum SO3 content in cement mortar. The effect of C3A on optimum SO3 content is about twice that of C4AF. The study also showed a trend of positive and important effect of the fineness of cement.