To objectively capture and understand tactical considerations in a race, the authors explored whether race-to-race variation of an athlete and the variation of competitors within a race could provide insight into how and when athletes modify their pacing decisions in response to other competitors.Lap times of elite 500-, 1000-, and 1500-m short-track speed-skating competitions from 2011 to 2016 (N = 6965 races) were collected. Log-transformed lap and finishing times were analyzed with mixed linear models. To determine within-athlete race-to-race variability, athlete identity (between-athletes differences) and the residual (within-athlete race-to-race variation) were added as random effects. To determine race variability, race identity (between-races differences) and the residual (within-race variation) were added as random effects. Separate analyses were performed for each event.Within-athlete race-to-race variability of the finishing times increased with prolonged distance of the event (500-m, CV = 1.6%; 1000-m, CV = 2.8%; 1500-m, CV = 4.1%), mainly due to higher within-athlete race-to-race variability in the initial phase of 1000-m (3.3-6.9%) and 1500-m competitions (8.7-12.2%). During these early stages, within-race variability is relatively low in 1000-m (1.1-1.4%) and 1500-m (1.3-2.8%) competitions.The present study demonstrated how analyses of athlete and race variability could provide insight into tactical pacing decisions in sports where finishing position is emphasized over finishing time. The high variability of short-track skaters is a result of the decision to alter initial pacing behavior based on the behavior of other competitors in their race, emphasizing the importance of athlete-environment interactions in the context of pacing.