PurposeLaboratory tests are commonly performed by cross‐country (XC) skiers due to the challenges of obtaining reliable performance indicators on snow. However, only a few studies have reported reliability data for ski‐specific test protocols. Therefore, this study examined the test–retest reliability of ski‐specific aerobic, sprint, and neuromuscular performance tests.MethodsThirty‐nine highly trained XC skiers (26 men and 13 women, age: 22 ± 4 years, V̇O2max: 70.1 ± 4.5 and 58.8 ± 4.4 mL·kg−1·min−1, respectively) performed two test trials within 6 days of a diagonal V̇O2max test, n = 27; skating graded exercise test to assess the second lactate threshold (LT2), n = 27; 24‐min double poling time trial (24‐min DP, n = 25), double poling sprint test (SprintDP1, n = 27), and 1‐min self‐paced skating sprint test (Sprint1‐min, n = 26) using roller skis on a treadmill, and an upper‐body strength test (UB‐ST, n = 27) to assess peak power (Ppeak) with light, medium, and heavy loads. For each test, the coefficient of variation (CV), intraclass correlation coefficient (ICC), and minimal detectable change (MDC) were calculated.ResultsV̇O2max demonstrated good‐to‐excellent reliability (CV = 1.4%; ICC = 0.99; MDC = 112 mL·min−1), whereas moderate‐to‐excellent reliability was found for LT2 (CV = 3.1%; ICC = 0.95). Performance during 24‐min DP, SprintDP1, and Sprint1‐min showed good‐to‐excellent reliability (CV = 1.0%–2.3%; ICC = 0.96–0.99). Absolute reliability for UB‐ST Ppeak was poor (CV = 4.9%–7.8%), while relative reliability was excellent (ICC = 0.93–0.97) across the loads.ConclusionIn highly trained XC skiers, sport‐specific aerobic and sprint performance tests demonstrated high test–retest reliability, while neuromuscular performance for the upper body was less reliable. Using the presented protocols, practitioners can assess within‐ and between‐season changes in relevant performance indicators.