Because of various conditional inspection data, some procedures have been reported for calculating the Markov transition probability for deterioration prediction. In this paper, it is tried to verify the property of those calculation procedures based on the given data in simple Markov process. In this paper, 3 proposed procedures are verified. Based on enough amount of data, those 3 procedures give almost equal transition probabilities. But, in the case that the numbers of data of each deterioration level have deviations, those 3 procedures give the transition probabilities which have differences not to be disregarded.