Minimizing life cycle cost (LCC) of bridge system while retaining healthy maintenance service level is a crucial task for bridge asset managers. Searching for optimized repair scenario for a bridge having different characteristics in different environments is a key to accomplish the task. This paper proposes a concept to generate all possible maintenance scenarios automatically during predicting degree of deterioration. The scenario that gives the lowest LCC with satisfied service level will be considered as an optimized repair scenario. The basic timing to repair is decided at the time just before deterioration state will be shifted to more severe level. However, there is no proof that repairing action in early point can give a better solution. Considering repairing timing at arbitrary time in scenario generation is an issue to discuss in this paper. Balance between computing time and effectiveness is also in discussion.