TRAILGENDER: A Sex-Disaggregated Database of Trail Running and Ultra-Trail Performances Worldwide (2010–2021)

TRAILGENDER is a comprehensive, open-science database documenting gender disparities in trail running and ultra-trail performance worldwide. The dataset covers 15,848 race editions held between 2010 and 2021 across 101 countries, aggregating performance records from three independent sources (Les Trails/TRAP, BjnNowak/Kaggle, and original web scraping). It includes individual-level data for 1,081,544 unique athletes and race-level metrics across nine standardised distance brackets (10K to 200K+/EXTREME). Key variables include the gender performance gap (GPG), computed as the relative difference between median female and male finish times; the female participation ratio by edition, country, and distance bracket; and a novel Gender Selection Index (ISG) measuring the gap between observed and expected female participation. The database is structured in 12 worksheets (Excel) with parallel CSV exports, a 121-variable data dictionary, 23 publication-quality figures, reproducible Python scripts, and full sensitivity analyses. TRAILGENDER supports reproducible research on gender equity in endurance sports, intersectional analyses (gender × age × nationality), sporting participation geography, and the socio-cultural dynamics of trail-running communities. It is designed as a FAIR-compliant research dataset for sports sociology, exercise science, and gender studies. The primary deposit is on Nakala/Huma-Num (DOI: 10.34847/nkl.cbbc2i86). This Zenodo record maximises international discoverability via OpenAIRE, DataCite, and Google Dataset Search.