Gait analysis in a laboratory may be expensive, time-consuming, and restricted to a controlled environment. Validated wearable technology may be an alternative to such analysis. However, wearable technologies should demonstrate reference values of a healthy population. Objective: To establish spatio-temporal gait reference values of an accelerometer (G-Walk) in a healthy Brazilian population. Methods: This is a cross-sectional study with 124 healthy subjects evaluated with G-Walk in the 6-minute and 10-meter walk tests (6MWT and 10MWT). Gait parameters of Velocity, Cadence, Distance, and gait symmetry variables were retrieved for analysis. Clinical and demographical characteristics were also collected and tested with simple linear regression as covariables of the gait characteristics. The bootstrapped 5th percentile of the gait parameter established the reference values. If a covariable influence was found, the reference values were established by subgroup analysis according to the covariable. Results: The study analyzed 114 subjects, mostly women (67.74%), aged 39.36 (SD 12.18). Height was a covariable of cadence for the 10MWT and cadence and stride length for the 6MWT. Age and sex combined were covariables of 6MWT velocity, and sex alone was a covariable of 6MWT. All reference values for symmetry were above 89%, velocity at the 10MWT was above 1.0m/s, and distance at the 6MWT was 354m and 359.5 for females and males, respectively. Conclusions: Our study generated reference values for spatio-temporal gait analysis with G-Walk of a population of a major urban area, considering the covariables of age, height, and sex.