In order to ensure the reliability and economy of isolated microgrid, it is important to set up a certain amount of operating reserve capability. The power fluctuation inside the microgrid mainly results from the uncontrollability of renewable distributed generation output, the failure outage of power generators and the inaccuracy of load prediction. This paper analyzes the probability and statistics pattern of the wind turbine output, photovoltaic power generation output and the short-term load prediction error, and then proposes a new optimizing strategy of reserve capacity based on the time-of-use price (TOU) to coordinate the reserve capacity between the demand side and the generation side. Aiming at minimizing the total procurement cost of operating reserve, under the chance constraint of meeting requirements of microgrid at a certain confidence level, a mathematical model of optimal reserve capacity is established. The mathematical model is solved by means of the hierarchical genetic algorithm (HGA) based on Monte-Carlo stochastic simulation. The effectiveness of the model is verified through the simulation of a practical microgrid. This work is supported by National Natural Science Foundation of China (No. 51277057).