Multi-centre gait biomechanics studies provide the opportunity to increase sample size and the confidence in results, yet differences between centres may introduce additional error. While previous investigations have compared gait biomechanics from different laboratories assessed by different raters, estimates of relative reliability, measurement error, and thresholds for real change are still unknown. These metrics are imperative to interpret multi-centre study results. Therefore, we examined the reliability of gait biomechanics assessed in two different laboratories, by two different raters, and using the same study sample. Twelve healthy participants underwent gait assessments by two raters in two laboratories at different institutions. Identical protocols were used to collect five walking trials per participant. Discrete data were examined for knee joint angles and moments, gait speed, and stride length. Reliability was assessed using the intraclass correlation coefficient (ICC), standardized error of the measurement (SEM), minimum detectable difference (MDD), and Bland and Altman plots. All spatiotemporal, joint angle, and joint moment measures had ICCs = 0.83–0.94, except for the knee adduction moment peak in late stance (ICC = 0.69 and 0.72). The knee adduction moment SEMs indicate that measurement errors due to the laboratory effect are between 0.19% and 0.31% body weight times height (0.03–0.05 Nm/kg). Meanwhile, measurement error for the knee flexion-extension angle is <2°. Our results are similar to previous test-retest reliability results from a single laboratory, and multiple laboratories, while adding previously unreported reliability metrics necessary for interpreting multi-centre study results.