The invention discloses a method for optimizing a power outage system recovery path based on a network flow theory. The method comprises the following steps: A) describing recovery of a power outage system as a single-source multi-sink directed network; B) analytically expressing a connectivity constraint in recovery path optimization based on the network flow theory; C) determining a line commissioning state; D) building a mixed integer line planning model for power outage system recovery path optimization; and E) performing solving by adopting CPLEX to obtain an optimal recovery path. Through adoption of the method, a non-linear model is transformed into the mixed integer line planning model, so that a rapid mixed integer model solving method can be adopted, and high optimizing efficiency and high optimizing result stability are achieved. The method can be applied to recovery path optimization of large-scale power outage systems, meets a requirement on calculation time, and has a good application prospect.