In sensed buildings, information related to occupant movement helps optimize important functionalities such as security enhancement, energy management, and caregiving. Typical sensing approaches for occupant tracking rely on mobile devices and cameras. These systems compromise the privacy of building occupants and inherently affect their behavior. Occupant detection and tracking using floor-vibration measurements that are induced by footsteps is a sensing method that is much less intrusive than use of cameras. Detecting the presence of occupants on a floor is challenging due to ambient noise that may mask footstep-induced floor vibrations. Additionally, spurious events such as door closing and falling objects may produce vibrations that are similar to footstep impacts. These events have to be detected and disregarded. Tracking occupants is complicated due to uncertainties associated with walking styles, walking speed, shoe type, health, and mood. Also, spatial variation in structural behavior of floor slabs adds ambiguity to the task of occupant tracking, which cannot be addressed using data-driven strategies alone. In this thesis, a framework for occupant detection and tracking is developed. In this framework, occupant detection is carried out based on signal information. This method outperforms existing threshold-based methods. Support-vector-machine classifiers, trained with time and frequency-domain features, successfully distinguish footsteps from spurious events and determine the number of occupants walking simultaneously. Inspired by pattern recognition using convolutional neural networks, the recognition of occupant characteristics is successfully carried out using footstep-induced floor vibrations. A model-based data-interpretation approach is used for occupant tracking. Structural-mechanics models are used to identify a population of possible occupant locations and trajectories. Up to two occupants can be tracked by accommodating systematic bias and uncertainties from sources such as modeling assumptions and variability in walking gaits. A zone-based approach to quantify footstep-contact dynamics is developed that improves the precision of results. In this approach, the floor-slab is divided into zones using knowledge of structural behavior. Within these zones, based on clustering of prior independent vibration measurements, severity levels of footstep impacts on the floor are determined. This helps reduce uncertainty related to walking-gait variability and improves the model of the footstep loading, thereby enhancing precision of occupant tracking. A hybrid framework for occupant detection and tracking that combines model-free approaches for occupancy detection with structural behavior models for tracking is developed and tested on four full-scale case studies. These studies successfully validate the utility of the framework for buildings having sparse sensor configurations that measure floor vibrations.