The invention relates to a dynamic spanning tree method for a topological incomplete measureable network. The method includes the steps of setting a starting threshold for a root node and low complexity node adding method during initialization, waiting for adding or withdrawing signals, going to the next step if a signal is detected, determining the type of the detected signal, if the signal type is node withdrawing, realizing withdrawing of the node from a tree, otherwise, determining whether the total number of the current tree nodes reaches the starting threshold of the low complexity node adding method, if yes, using the low complexity node adding method to add the node in the tree, otherwise, starting the process of adding nodes to the tree by using a high precision node adding method, verifying whether the topological information is completely measurable during the node adding process, realizing process of adding nodes to the tree by using the low complexity nodes adding method once the topological information is incompletely measurable, otherwise, completing the node adding process by using the high precision node adding method.