RISE Fire Technology has replaced the use of the desiccant DRIERITE for drying smoke gases with Peltier and high compression cooling for the Servomex ServoPro 4100 O2 and CO2 analyser used for the medium-scale cable testing according to EN 50399. The replacement is made mainly for work environment improvement purposes. A prerequisite for measuring correctly is that the concentration of water vapor prior to the analysers is low, and most important, has a low variation. An increasing amount of water vapor due to a change in drying efficiency, will proportionally dilute the species concentration in the gases introduced into the analyser, thus giving a measurements error of the HRR (Heat Release Rate). Prior to the change of drying method, tests were made to ensure that the measurements using the new drying unit were correct and according to standard. The following tests were performed: 1. Sampling ambient air to evaluate the drift in humidity, O2 and CO2 over time. 2. Sampling ambient air during changes between the two Peltier units of the new drying unit, evaluating instantaneous humidity changes. 3. Sampling the effluents from a 20.5 kW propane burner directed against a wet wooden chip board to evaluate the new drying unit’s ability to dry out high concentrations of water vapour. 4. Sampling ambient air during different seasons where the RH in the surrounding air differs, evaluating the drifts at different moisture levels. 5. Conducting all calibrations according to EN 50399 which affect O2 and CO2 and HRR. Performed with DRIERITE as well as with the new drying unit, comparing them and assuring that the calibration criteria are passed. The evaluation of the Peltier and high compression drying unit gave the following results: 1. A small negative drift was seen in the H2O level reaching the O2 and CO2 analysers over time. There is a reverse linear correlation with the O2 reading which rises when the H2O level drops. By calculating the influence on HRR and using RMS over 30 minutes, a drift in HRR of 0.36 kW can be seen. The corresponding THR during a test can be calculated to 0.2 MJ. These levels of drifts in HRR and THR can be regarded as acceptable compared to other uncertainties in the HRR and THR measurements. 2. It was found that there are instantaneous humidity changes that have an impact on the analysers when changing between the two Peltier coolers. The corresponding HRR changes are calculated to 1.5 kW, which cannot be regarded as acceptable. Thus, a change of Peltier coolers should not be accepted during a test, nor within 10 minutes prior to or after a test. 3. The results of the fire test with the chip board confirm the ability of the new cooling unit’s technique to dehumidify combustion gases to an extent that can be regarded as a worst case. The levels were also smooth with a low drift. 4. Tests that were performed during different seasons showed that the absolute drift may be higher during seasons with a higher RH. However, still at acceptable levels. 5. All calibrations were approved with a margin, both for the DRIERITE as well as for the new drying unit. The conclusion is that the new drying unit solution works well from a measurement uncertainty point of view. However, good routines are needed to ensure that testing does not take place when this drying solution changes the Peltier element cooler.