The ability to ambulate within one’s environment is an important skill for everyday life, however, traditional methods of assessing gait involve walking in controlled settings that may not represent performance in real-world environments. Laboratory and clinically-based quantitative gait assessments are often conducted in a sterile and uniform environment and treadmill walking assessments, used to collect a large amount of walking data, are conducted with an externally driven speed and fixed environment. As such, gait demonstrated in these assessments may not reflect gait performance in the real-world environment. PURPOSE: To compare gait behavior among overground indoor, outdoor, and treadmill walking in healthy adults. METHODS: 16 healthy young adults (5M, 11F, 21±2yrs) performed three walking trials (500m each) at a self-selected pace: indoors around a gymnasium (IN), outdoors along a sidewalk (OUT), and on a treadmill (TM). Data were collected using wearable sensors and the following variables were calculated for each trial: cadence, gait velocity, double support %, step time, stride length. Two repeated-measures MANOVAs were used to compare mean values and variability (standard deviation (SD)) across conditions. RESULTS: Significant differences were found across the conditions for both mean values (Λ*=.038, F(10,6)= 15.18, p=.002), and variability (Λ*=.037, F(10,6)= 15.41, p=.002). Univariate tests showed differences in mean values between all three conditions for cadence, gait velocity, step time, and stride length (p<.003), and greater double support time in the TM condition than the IN or OUT conditions (p=.006). Univariate tests for variability measures failed to detect statistically significant differences (p>.01). The TM condition had a lower gait velocity SD than IN or OUT (p=.001). CONCLUSION: Healthy young adults adopt different walking strategies while walking indoors, outdoors, and on a treadmill.