Background There is no standardized method for the assessment of physical activity (PA). Therefore it is important to investigate the validity and comparability of different measures. The International Physical Activity Questionnaire (IPAQ) has been developed as an instrument for cross-national assessment of PA and has been validated in 12 countries. These instruments have acceptable measurement properties for monitoring population levels of PA among 18–65 year-old adults in diverse settings. However, there are some concerns that IPAQ may over-report PA. The purpose of this study is to evaluate the reliability and validity of IPAQ, short version, last 7-days in the Nord-Trøndelag Health Study (HUNT) population of men. Methods The questionnaire was administered twice to a random sample of 108 men aged 20–39 and validity by comparing results with VO2max and ActiReg, an instrument that measures PA and energy expenditure (EE). ActiReg discriminates between the body positions: stand, sit, bend forward and lie and also registers if there is motion or not in each of them or both. Results Our results for reliability of the IPAQ short version were good for vigorous and fair for moderate activities. Intraclass correlations ranged from a low of 0.30 for moderate activity hours, to a high of 0.80 for sitting hours. Concerning validity, our results suggest that total IPAQ vigorous PA was a moderately good measure of vigorous activity, having moderately strong, significant correlations with VO2max, r = 0.41 (p ≤ 0.01), but correlated not with metabolic equivalent (METs) values of 6 or more measured with ActiReg. Only total IPAQ walking was fair correlated with METs 1–3 and METs 3–6, respectively r = -0.27 and 0.26 (p ≤ 0.05). The index for IPAQ sitting hours per week was moderate correlated with METs values of 1–3 and negatively correlated with METs values of 3–6. Classification of PA in three levels (low, moderate and high) correlated also most strongly with VO2max (0.31 p ≤ 0.01) and METs 3–6 and METs 1–3 from ActiReg (r = 0.32 and -0.31, p ≤ 0.01). Classification of BMI in three levels (normal, overweight and obese) correlated most strongly negative with VO2max (-0.42 p ≤ 0.01) and MJ from ActiReg (r = 0.31 p ≤ 0.01). Conclusion Our results indicate that IPAQ short version for men has acceptable reliability and criterion validity for vigorous activity and sitting. Walking has moderate reliability. Only the IPAQ for walking had a fair correlation with METs 6+. The questions about moderate activity had fair reproducibility and correlated poorly with most comparison measures.