Currently, studies on the system-of-systems (SoS) capability evaluation index system generation mechanism are scant, and the indicators are mostly determined by merely qualitative expert experience. This paper tries to explore a quantitative method for constructing the search and rescue system-of-systems (SARSoS) capability evaluation index system based on the SoS network architecture analysis, considering the impact from the SoS activity, organization, equipment, project, technology, and their relationships constraint on the capability generation. The four-layers weighted supernetwork model of the SARSoS is first established to describe the influence of SoS network architecture to the capability. Then, the index system construction method is proposed through the indicator importance sort algorithm (IISA). Finally, a specific case study of the SARSoS capability evaluation index system construction is demonstrated based on the proposed method. This paper provides a viable way to construct the index system based on the SoS architecture, which can select and integrate the indicators from multiple perspectives, and help to discover the SoS weak points and relationships in the future capability evaluation work.INDEX TERMS Search and rescue system-of-systems, capability evaluation, index system construction, weighted supernetwork model, indicator importance sort algorithm.