Current infrastructure is designed and built such that it must resistall possible loads. This leads to overdesigned structures that are inefficient in terms energy and cost. A structure that can self-identify damage, adapt, and learn for future events results from research intothe emerging field of intelligent infrastructure and structural health monitoring.Two halves of a “hollow-rope” tensegrity structure deploy from supports to join at midspan by controlling the length of active cables on each half of the structure. These active cables are continuous throughthe length of the half-structure, guided by intermediary jointswhere cables slide. Springs along the circumference of the structure facilitate deployment due to increasing the diameter of the structure when folding and subsequent decreasing during deployment. Although previous work has addressed damage locationand mitigationof ruptured cableswhen the cables are the load-critical elements of the structure, this work hasnot studiedthe classification of the type of element that is damaged, element locationand damage mitigation.This paper presents work on element classification, detection,and location of damaged elements in a deployable tensegrity footbridge. The footbridge isstudied through monitoring dynamic behavior. Displacement and strain values are measured before, during, and after cable breakage. Natural frequencies inhealthy and damaged states are compared. Free-vibration dynamic behavior of the tensegrity structure are characterized for two situations,deployment and in-service. Examination of ambient vibrations for the half structure and forced vibrations for the full structure successfully led to detection of ruptured cables. Correlation methods using strain measurements also successfullydetect and locate a ruptured cable. Detection of abuckled strut and aruptured cableis successful by observing differences of natural frequencies between healthy and damaged states. Location of a damaged element is successful using nodal-position measurements through excluding possible damage scenarios and using strain measurements to identify elements of significant changes in eigenvector coefficients using principal component analysis. Therefore, excluding scenarios from a population for damageidentification is effective for highly-coupled structures that are capable of large shape changes. These methods reveal the potential for damage identification of complex sensed structures.Classification and location of a damaged element on a complex near-full-scale structure is successful using nodal position measurements through excluding possible damage cases and using strain measurements to identify elements of significant changes in eigenvector coefficients using principal component analysis. Implementing error-domain model falsification to exclude possible scenarios for location of damaged elements successfully reduced the number of probable casesof damage location. Paterns of influence from damaged cables and struts are useful to classify the type of element that is damaged. Therefore, the methodology involving error-domain model falsification (EDMF) for damage location is useful for closely-coupled structures that are capable of large shape changes.