Objectives: To examine the within- and between-player variability of physical performance and player match loads in professional rugby union. Design: A single cohort, observational study. Methods: Physical match performance data were collected from 28 male, professional, English Championship players over 15 competitive matches. Using microsensors, the variables selected for analysis were total distance (TD), low-speed running distance (LSR), high-speed running distance (HSR), very high-speed running distance (VHSR), total impacts (TI), repeated high-intensity efforts (RHIE), body load (PlayerLoad™; PL), and low velocity (<7.2 km·h-1) body load (PLSLOW). Ratings of perceived exertion (RPE) represented match internal loads. Variability was quantified using the coefficient of variation (CV), with the meaningful interpretation of change in physical performance and match loads calculated using magnitude-based inferences. Results: We found large between-match (within-player) variation for HSR (27.6%; ±90% confidence limits 6.9% [forwards], 20.1%; ±4.1% [backs]), VHSR (68%; ±19%, 34.1%; ±7.5%), TI (24.0%; ±5.9%, 36.4%; ±7.9%) and RHIE (18.7%; ±4.4%, 39.5%; ±8.8%), with moderate variability for match RPE (8.2%; ±1.8%, 10.8%; ±2.1%), PL (7.3%; ±1.7%, 10.0%; ±2.0%) and PLSLOW (8.9%; ±2.0%, 10.7%; ±2.1%). Threshold values for likely substantial between-match changes in high-intensity physical performance measures ranged from 21–76%, and were ~10% for match RPE, PL and PLSLOW. Conclusions: Within- and between-player variability of high-intensity activity in professional rugby union is large, yet RPE, PL and PLSLOW appear more stable by comparison and may be interpreted with greater accuracy.