In maintenance and rehabilitation (M&R) decision-making for infrastructures, there are many cases that their record unit of deterioration conditions by inspection works (inspection unit) and their decision making unit of M&R works (M&R unit) are different. In this paper, an aggregative deterioration process model is proposed to evaluate deterioration conditions in the M&R unit based on multiple information of deterioration condition states obtained in inspection units. The authors a) express infrastructure deterioration processes in inspection units by a Markov chain model and b) formulate a stochastic process model expresses occurrence probabilities of condition state patterns in the M&R unit with the use of multinomial distributions as the aggregative deterioration process model. Then, a risk indicator in M&R units is defined using the proposed method without omission of information of inspection units. The risk indicator enables a relative evaluation across M&R units. Lastly, the effectiveness of the proposed methodology can be discussed through the empirical study targeting RC slabs on actual expressway viaducts.