Substructure identification methods may be used for structural health monitoring to isolate a small section of a structure so that damage may be identified within the substructure area and ignored outside of the substructure area. An application of the Remote Sensing System (RSS) is developed to compute the frequency response functions of an isolated substructure under ambient loading conditions. This isolation technique is shown to be equivalent to the Craig-Bampton substructural representation when loading is collocated with the interfaces of the substructure and all interface degrees-of-freedom are measured. The measurements at the interfaces and inside the isolated substructure are treated as inputs and outputs, respectively, for system identification. The identified RSS frequency response functions characterize the local modal characteristics of the substructure as though the interfaces are fixed. The frequency response assurance criterion (FRAC) is used to establish the presence of damage in the structure through comparison with a baseline, undamaged, RSS frequency response function. This frequency-domain damage identification is shown to be successful at detecting stiffness changes within the substructure and neglecting stiffness changes outside the substructure, as illustrated computationally using a spring-mass problem.