Purpose <p>In long-track speed skating, specifically team events, drafting is crucial for success. Although skaters and coaches are aware of the benefits of drafting, the lack of a practical protocol to assess on-ice drafting skills inhibits skill-specific training. The aim of this research is to develop a practical protocol to assess on-ice drafting skills in skating (DRAFT) and evaluate its reliability and validity.</p> Method <p>A video-based protocol was developed to assess three key drafting skills: distance between the skaters, following in a straight line, and takeover in the corner. This research assessed the drafting skills of 29 junior speed skaters (14 male, 15 female; 16 ± 1&#xa0;years old; 7.4 ± 1.6&#xa0;years skating experience). Intra- and inter-rater reliability (relative and absolute) and concurrent validity were established. Discriminative validity was tested by grouping skaters (good, moderate, poor) via expert ratings and analyzed using one-way ANOVA (<i>p</i> &lt; .05).</p> Results <p>Relative reliability was excellent (ICCs &gt; .91, <i>p</i> &lt; .01) in all drafting skills. Absolute reliability was good, except for intra-rater reliability in distance between skaters (95% CI [0.02, 0.06], <i>t</i>(21) = 5.25, <i>p</i> &lt; .001, <i>d</i> = 1.12). Concurrent validity was supported for distance between the skaters (<i>ρ</i> = −0.66, <i>p</i> = .01) and following in a straight line (<i>ρ</i> = 0.89, <i>p</i> &lt; .001), but not for takeover <i>(ρ</i> = –.18,<i> p</i> = .52). These drafting skills showed a significant between-group difference, with better scores for good drafters (<i>p</i> &lt; .05), no such differences were found for takeover (<i>p</i> = .63).</p> Conclusions <p>The innovative and practical protocol DRAFT can be used for analyzing drafting skills on straightaways, while more research is needed on the corners. The protocol enables targeted training to optimize performance in long-track speed skating.</p>