Ambulatory health monitoring has been demonstrated to provide significant benefits to patients in a number of studies. The biggest promise for improved results for patients lies in the long-term, ambulatory monitoring of patients with chronic illnesses to help doctors and patients identify health issues before they reach a crisis stage. This project pursues the design of garments for ambulatory health monitoring that have the look and feel of every day clothing, together with an approach to monitoring that removes some of the barriers to patient compliance by automatically annotating physiological data with activities and motions, collecting data only during specific conditions, and having minimal impact on daily routine. The intellectual merits lie in the algorithms, design methodologies, and evaluation methodologies that enable wear-and-forget garments for ambulatory health monitoring. The approach developed by this project uses a computationally intensive pose estimation algorithm that automatically adapts a more computationally efficient algorithm for activity classification. This project addresses the challenges of providing garments that look and feel like everyday clothing by developing design and evaluation methodologies for incorporating fiber-based sensors into garments. These have significant advantages over discrete sensors by being woven or sewn into the fabric, covering much larger areas of a garment, and draping naturally. The broader impacts of the research lie in the potential to remove barriers that prevent the effective use of ambulatory health monitoring to improve the quality of life for patients and their families.