Depending on the phase of the menstrual cycle, different values of running kinematic variables can be obtained. The aim of this study is to analyze whether there are changes in the kinematic variables in running throughout the menstrual cycle and to relate them to running performance and injury prevention. Eight regular female runners and triathletes performed a maximal treadmill test, as well as a submaximal test (6′ stages at 50%, 60% and 80% of maximal aerobic speed) in each of the phases of the menstrual cycle: menstruation phase (day 2.4 ± 0.7), follicular phase (day 10.4 ± 2.2) and luteal phase (day 21.8 ± 2.1). Running dynamics were measured using RunScribe. For parametric data, a general linear model of repeated measures was applied, with two intrasubject independent variables, menstrual cycle phases (with three levels: Menstruation, Follicular, and Luteal) and running intensity (with four levels relative to the maximum speed reached in the test: 100%, 80%, 60%, and 50%). For variables with non-normal distributions, Friedman tests were performed with Wilcoxon post-tests adjusted for significance according to Bonferroni. The maximum stance velocity from foot strike to the point of maximum pronation (°/s) was higher in the menstruation phase than in the follicular and luteal phases (p = 0.008), the step rate (s/min) was higher in the follicular phase than in the menstruation and luteal phases (p = 0.049), the vertical velocity (m/s) was lower in the follicular phase than in the menstruation (p = 0.004) and luteal phases (p = 0.003), and the contact time (ms) was lower in the luteal phase than in the menstruation and follicular phases. These results suggest that training at high intensities could be a factor in greater risk of injury in female athletes, especially in the menstruation phase, finding in the luteal phase and at an intensity of 80% a greater efficiency in the running.