Compressive strength in concrete increases with time. Regression analysis with time is conventionally performed for strength evaluation and prediction. In this study, hydrate amount is assumed as a function of hydration rate and porosity, and modeling on compressive strength is carried out considering decreasing porosity with time, which does not need the regression analysis with time. For twenty one mix proportions of HPC (High Performance Concrete), DUCOM (FE program) which can simulate the behavior in early aged concrete is utilized, and porosity from each mix proportions is obtained with time. For HPC with OPC (Ordinary Portland Cement) concrete, modeling on compressive strength is performed considering hydration rate, unit content of cement, and porosity with time. For HPC with mineral admixtures, a long-term parameter which can handle long-term strength development is additionally considered. From the comparison with the previous test results, the applicability of the proposed model is verified.