ABSTRACT: The use of physical tests to profile physical capabilities, and provide training direction to athletes is common practice. Likewise, in professional team sports, notational analysis codes the key contributions of each player during competition. Limited studies have however investigated relationships between physical capabilities and key performance indicators (KPIs) of rugby union match-play. Elite professional players, categorised as forwards (n = 15) or backs (n = 14), from an international rugby union squad (n = 29) undertook assessments of isometric mid-thigh pull (IMTP), bilateral and unilateral countermovement jumps (CMJ) and drop jumps (DJ; from 40 and 20 cm, respectively), and assessment of acceleration (10 m), a 5 m weighted sled drive, and a Yo-Yo intermittent recovery test level 1 (Yo-Yo IRTL1). Game statistics of the same players from 92 matches (~23 matches per player) during the 2014-15 season were analysed for effort and performance-based metrics. For forwards, Yo-Yo IRTL1 correlated significantly with; number of tackles made (r = 0.717), first three players at a ruck in both attack (r = 0.568) and defence (r = 0.581), number of effective rucks (r = 0.630), total possessions (r = 0.522), passes made (r = 0.651), percentage of carries over the gainline (r = 0.610), effective ruck success (r = 0.600), tackle success (r = 0.540), and the number of turnovers made (r = 0.518). Drop jump performance in forwards was associated with; the number of clean breaks (r = 0.558), dominant collisions (r = 0.589), and offloads (r = 0.594). For backs, the sled-drive test correlated with; number of carries (r = -0.751), first three players at an attacking ruck (r = -0.613), effective attacking rucks (r = -0.584), number of dominant collisions (r = -0.792) and offloads (r = -0.814). Likewise, for backs, IMTP peak force was related to; the number of possessions (r = 0.793), passes made (r = 0.792), effective attacking ruck percentage (r = 0.628), and the number of offloads (r = 0.621) whilst relative peak force correlated with; the percentage of carries over the gainline (r = 0.533), percent tackle success (r = 0.603) and effective attacking ruck percentage (r = 0.584). Regression analyses highlighted that only a small number of variables (i.e., carries, tackles, attacking and defensive first three at ruck) returned practically achievable changes (<20%) in physical qualities. In spite of this, and while leaving scope identification of further physical and/or performance predictors, greater strength, power and intermittent running performance were positively related to match-derived KPIs during competition. This may provide a basis for better integrating the strategies used by physical and technical performance-focused coaching staff to improve key performance indicators, and thus match performance, of rugby union players.