Thirty-eight competitive cross-country skiers were divided into three groups to assess the reliability and validity of a new double poling ergometer. Group A (n = 22) performed two maximal 60-s tests, Group B (n = 8) repeated peak oxygen uptake tests on the double poling ergometer, and Group C (n = 8) performed a maximal 6-min test on the double poling ergometer and a double poling time-trial on snow. The correlation between the power calculated at the flywheel and the power applied at the base of the poles was r = 0.99 (P < 0.05). The power at the poles was 50 - 70% higher than that at the flywheel. There was a high test - retest reliability in the two 60-s power output tests (coefficient of variation = 3.0%) and no significant difference in peak oxygen uptake in the two 6-min all-out tests (coefficient of variation = 2.4%). There was a strong correlation between the absolute (W) and relative power (W . kg(-1)) output in the 6-min double poling ergometer test and the double poling performance on snow (r = 0.86 and 0.89 respectively; both P < 0.05). In conclusion, our results show that the double poling ergometer has both high reliability and validity. However, the power calculated at the flywheel underestimated the total power produced and needs to be corrected for in ergonomic estimations.