The aim of this study was to evaluate two versions of the Oxycon Mobile portable metabolic system (OMPS1 and OMPS2) in a wide range of oxygen uptake, using the Douglas bag method (DBM) as criterion method. The metabolic variables VO2, VCO2 respiratory exchange ratio and VE were measured during submaximal and maximal cycle ergometer exercise with sedentary, moderately trained individuals and athletes as participants. Test-retest reliability was investigated using the OMPS1. The coefficients of variation varied between 2 and 7% for the metabolic parameters measured at different work rates and resembled those obtained with the DBM. With the OMPS1, systematic errors were found in the determination of VO2 and VCO2 At submaximal work rates VO2 was 6-14% and VCO2 5-9% higher than with the DBM. At VO2(max) both VO2 and VCO2 were slightly lower as compared to DBM (-4.1 and -2.8% respectively). With OMPS2, VO2 was determined accurately within a wide measurement range (about 1-5.5 L min(-1)), while VCO2 was overestimated (3-7%). VE was accurate at submaximal work rates with both OMPS1 and OMPS2, whereas underestimations (4-8%) were noted at VO2(max). The present study is the first to demonstrate that a wide range of VO2 can be measured accurately with the Oxycon Mobile portable metabolic system (second generation). Future investigations are suggested to clarify reasons for the small errors noted for VE and VCO2 versus the Douglas bag measurements, and also to gain knowledge of the performance of the device under applied and non-laboratory conditions.