Sulfate erosion is a major cause of concrete durability deteriorations, especially for the service tunnels which are in sulfate erosion for a long time. Accurately predicting the concrete damage failure under sulfate erosion has been a challenging problem in the evaluation and maintenance of concrete structures. Here we design the dry-wet cycle test of service tunnel concrete under sulfate erosion, and analyze the Elastic relative dynamic modulus(Erd) and mass under 35 times cycle periods. Then we develop an autoregressive integrated moving average (ARIMA) prediction model linking damage failure to Erd and mass. The results show that the deterioration of concrete first increased and then decreased with an extension of the dry–wet cycle period. Moreover, based on a finite set of training data, the proposed prediction approach shows high accuracy for the changes of concrete damage failure parameters in or out of the training dataset. The ARIMA method is proven to be feasible and efficient for predicting the concrete damage failure of service tunnel under sulfate erosion for a long time.