Coordination variability (CV) is commonly analyzed to understand dynamical qualities of human locomotion. The purpose of this study was to develop guidelines for the number of trials required to inform the calculation of a stable mean lower limb CV during overground locomotion. Three-dimensional lower limb kinematics were captured for 10 recreational runners performing 20 trials each of preferred and fixed speed walking and running. Stance phase CV was calculated for 9 segment and joint couplings using a modified vector coding technique. The number of trials required to achieve a CV mean within 10% of 20 strides average was determined for each coupling and individual. The statistical outputs of mode (walking vs running) and speed (preferred vs fixed) were compared when informed by differing numbers of trials. A minimum of 11 trials were required for stable mean stance phase CV. With fewer than 11 trials, CV was underestimated and led to an oversight of significant differences between mode and speed. Future overground locomotion CV research in healthy populations using a vector coding approach should use 11 trials as a standard minimum. Researchers should be aware of the notable consequences of an insufficient number of trials for overall study findings.