Although previous studies have examined how the triathlon segments influence overall performance, how both segments and transitions differ by sex and across short-course distances, including the Super-Sprint, remains largely unknown. This study analysed how triathlon segments impacted overall performance of elite short-course triathletes, considering sex, race distance, and the performance level. Results from 113 World Triathlon Series (2017–2024), including Super-Sprint, Sprint, and Olympic races, were analysed for elite female and male triathletes (n = 5051). Split times were normalised to a 0–100 score and analysed using general linear models and decision tree analyses. Running performance showed the strongest associations across all distances, with β values ranging from 0.13 to 0.39 for males and 0.07 to 0.23 for females, for Super-Sprint to Olympic races, respectively. Running also showed the highest contribution to prediction capacity for Top-10 finishers. Although cycling's predictive capacity increased with distance, it remained lower than swimming for female athletes. Transition times, especially transition 1 (T1), had a notable influence on final positions in Sprint and Super-Sprint events, particularly among females (β = 0.13 and 0.07, respectively), but had minimal impact in the Olympic distance. Transition 2 (T2) also contributed to Sprint performance in both sexes (β = 0.05). Our findings reveal that, as competition distance increases, cycling and running become more crucial, with running being the most decisive factor for both sexes. T1 performance is more influential for female triathletes, especially in shorter distances, whereas swimming has minimal impact on overall results, except for Top-10 female finishers.