This study aimed to identify the aerobic and anaerobic metabolic performance capacities contributing to Yo-Yo Intermittent Recovery level 2 (Yo-Yo IR2) test performance. Nineteen recreational Australian footballers completed a Yo-Yo IR2 test, and on another day a treadmill peak oxygen uptake (VO_2peak) and maximal accumulated oxygen deficit test in a randomised counter-balanced order. The maximal accumulated oxygen deficit (MAOD) procedures included 5 × 5 min sub-maximal continuous runs at progressively higher speeds whilst VO₂ was recorded; thereafter, speed was incrementally increased to elicit VO_2peak. After 35 min of rest, participants ran at a speed equivalent to 115% of VO_2peak until exhaustion, at which point expired air was collected to determine maximal accumulated oxygen deficit. Relationships between variables were assessed using Pearson’s correlation and partial correlations. Maximum aerobic speed, relative intensity, and VO_2peak were significantly correlated with Yo-Yo IR2 performance. High Yo-Yo IR2 performers also had higher MAS, relative intensity, and VO_2peak levels. However, when higher maximum aerobic speed, relative intensity, and VO_2peak were controlled for each other and analysed independently, neither maximal aerobic speed nor VO_2peak correlated with Yo-Yo IR2 performance. Yo-Yo IR2 performance is the result of a complex interaction between several variables. Training programs should primarily focus on improving VO_2peak, maximal aerobic speed, and relative intensity to optimize Yo-Yo IR2 test performance.

Objectives: This study investigated the reliability and criterion validity of the Concept 2 SkiErg™ to assess 1,000-m on-snow, time trial performance using the classical double poling technique.Methods: Ten athletes (5 males and females) from a National cross-country ski team participated in the study and completed a 1,000-m time trial on snow, as well as two 1,000-m time trials on the Concept 2 SkiErg™ in a temperature-controlled room during a 4-day training camp.Results: There was a significant decrease in time from test 1 to test 2 of 4.87 s [238.3 ± 26.1 vs. 233.4 ± 23.9 s; 95% limits of agreement (LoA): −5.5, 15.3]. The Concept 2 SkiErg™ time-trial had a coefficient of variation (CV) of 1.6% and the standard error of measurement was 3.8 s. When compared to the on-snow time-trial, the Concept 2 SkiErg™ time-trial demonstrated a mean bias of 20.7 s (95% LoA; 11.6, 29.8) and the concordance correlation coefficient was 0.72.Conclusion: The Concept 2 SkiErg™ demonstrated excellent single-trial reliability. However, there was significant proportional bias in the Concept 2 SkiErg™ relative to the on-snow test and agreement between the two was relatively poor. Research using a larger sample and different trial durations is required to further validate the Concept 2 SkiErg™ for cross-country skiing performance testing.