The evaluation of global chemistry-climate/transport models in the upper troposphere – lower stratosphere (UTLS) is a important step towards a better understanding of the chemical composition near the tropopause, and therefore towards a more accurate assessment of the impact of NO_x emissions in this region of the atmosphere, notably by subsonic aviation. For this purpose, the current study focuses on an evaluation of long-term simulations from five global models based on in-situ measurements on board passenger aircraft (IAGOS). Most simulations span over the 1995–2017 time period, and follow a common protocol among the models. The assessment focuses on climatological averages of ozone (O₃), water vapour (H₂O), carbon monoxide (CO), and reactive nitrogen compounds (NO_y). In the extra-tropics, the models reproduce the seasonality of ozone, water vapour and NO_y in both the upper troposphere (UT) and the lowermost stratosphere (LS), but none of them reproduces the CO springtime maximum in the UT. The tropospheric tracers (CO and H₂O) tend to be underestimated by the models, which is consistent with an overestimation of the cross-tropopause exchange, but does not exclude other factors as an underestimation of CO emissions, an underestimation of transport from the surface, or an overestimated CO oxidation by the hydroxyl radical (OH). Ozone is systematically overestimated in the UT by most models, and the NO_x background appears as the main contributor to the ozone variability across the models. The partitioning between NO_y species changes drastically across the models, and acts as a source of uncertainty on the NO_x mixing ratio and on its impacts on atmospheric composition and particularly on the response to aviation NO_x emissions. However, independently on the mean biases, we highlight some well-reproduced geographical and seasonal distributions, as the ITCZ seasonal shifts above Africa, the upper-tropospheric H₂O maximum in the Asian summer monsoon, and the extratropical ozone (H₂O) in the LS (UT) that show a high correlation with the observations. These features are encouraging regarding the simulated dynamics in both the troposphere and the stratosphere. The current study confirms the importance of an accurate separation between the UT and LS using a dynamical tracer for model results evaluation but also for model intercomparisons.