In the current study, we examined whether sensorimotor information stored in short-term memory may influence the temporal features between overt and covert execution of human locomotor movements and, furthermore, to examine to what extent such influence may depend on the ongoing gait activity. The subjects (n=20) who participated in the experiment were separated in two groups and instructed to walk (overt execution) or imagine walking (covert execution) along three locomotor paths: horizontal, uphill and downhill. The subjects of the first group, labeled in block, performed all the covert trials before executing the corresponding overt trials, while the subjects of the second group, labeled in serial, alternated between overt and covert movements. The overt and covert durations were recorded by means of an electronic stopwatch. We found high temporal similarities between overt and covert execution for both groups and for all the locomotor paths. Nevertheless, the execution of imagined movements was more variable compared to their actual counterparts. In addition, timing variability of covert movements was smaller for the group in serial compared to the group in block. This decrease in timing variability was larger for the horizontal walking compared to uphill or downhill locomotion. Furthermore, linear regression analysis performed on the data of the group in serial showed close temporal relationships between each successive overt and covert execution. These relationships were stronger when walking along the horizontal vs. sloped paths. Therefore, our results on timing variability provide evidence for gait-dependent motor memory facilitation in covert movement execution. We propose that gait-dependent contribution of sensory feedback information to movement regulation may account for these findings