<h3>Research Objectives</h3> To investigate and compare the biomechanical characteristics and cardiopulmonary metabolic energy efficiency of gait between the young-old and the old-old community dwelling people. <h3>Design</h3> Cross-sectional study. <h3>Setting</h3> Clinical experimental study. <h3>Participants</h3> The elderly participants (N=78) were classified into the young-old (age range=65-74, mean age=70.2 ± 2.8 years, n=44) and the old-old (age range=75-84, mean age=77.9 ± 2.2 years, n=34) groups. <h3>Interventions</h3> Not applicable. <h3>Main Outcome Measures</h3> Spatiotemporal parameters (gait speed, cadence, step length, step width, and stride length), kinematics (angle of hip, knee, and ankle joint), kinetics (moment and power of hip, knee, and ankle joint), and cardiopulmonary metabolic energy efficiency during gait. <h3>Results</h3> Spatiotemporal gait parameters (gait speed, step length, and stride length) were significantly decreased in the old-old compared to the young-old group (p < 0.01). Peak ankle dorsiflexion angle, peak moment and peak power of knee and hip joints during gait were also significantly decreased in the old-old compared to the young-old group (p < 0.05, p < 0.01). On the other hand, net cardiopulmonary metabolic energy cost showed no significant difference between groups. <h3>Conclusions</h3> The old-old and young-old people showed significantly different gait characteristics, which suggested a need of customized intervention strategy in two elderly groups for preventing age-related hazardous condition such as fall. <h3>Author(s) Disclosures</h3> Eunhye Chung1,2, Su-Hyun Lee2, and Yun-Hee Kim1,2,3 1. Department of Medical Device Management & Research, SAIHST, Sungkyunkwan University, SAIHST, Seoul, Republic of Korea 2. Department of Physical and Rehabilitation Medicine, Center for Prevention and Rehabilitation, Heart Vascular Stroke Institute, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea 3. Department of Health Sciences and Technology, Department of Digital Health, Sungkyunkwan University, SAIHST, Seoul, Republic of Korea