The initiation and growth process of cyclic ice body in porous systems are affected by thermo-physical and mass transport properties as well as by gradients of temperature and chemical potentials. Furthermore, the diffusivity of deicing chemicals reaches a significantly higher value under cyclic freeze-haw conditions. Moreover, disintegration of concrete structures is aggravated by marine environments, higher altitudes, and northern areas. A serious concern for concrete engineers is that the property of cyclic freeze-haw with crack growth and the deterioration, caused by accumulated damages hard to be identified by testing. The input variables employed in the analytical predictions are selected as random variables for the limit state functions, followed by reliability analyses. For this purpose, a linear adaptive weighted response surface method (LAW-RSM) has been developed. The verification exhibits a reasonably good correlation with the exact solution of MCS. The important parameters for cyclic freeze-haw-deterioration of concrete structures as water to cement ratio, entrained air pores, and the number of cycles of freezing and thawing are used to construct the limit state function of LAW-RSM. The regression equation fitted to the important deterioration criteria such as accumulated plastic deformation, relative dynamic modulus and equivalent plastic deformations served as the probabilistic evaluation of the performance to resist the structural degradation. The composed improved response surfaces could be utilized to predict the equivalent plastic strains, relative dynamic modulus and residual strains during the cycles of freeze-haw with the probability of failure as changing input random variables. Hence, it is possible to evaluate the life cycle management of concrete structures by the proposed prediction method in consideration of the accumulated damage due to cyclic freeze-thaw.