One of the most fundamental questions in Atmospheric Chemistry for the beginning of the 21st Century is how the observed growth in anthropogenic emissions has the potential to affect the processes that occur much higher up in the stratosphere, especially the influence on the recovery time needed for the ozone layer as a consequence of the Montreal protocol. Moreover, future predictions suggest that the dominance of chlorine radicals and oxides in depleting stratospheric ozone will diminish by 2050, with bromine possibly becoming more and more influential. The subsequent transport of trace gases through the Upper Troposphere/Lower Stratosphere (UT/LS) which is simulated in any large-scale chemical model therefore needs to be quantified in order to assess whether such models have the ability to simulate realistic transport. One method for determining this is to use chemically passive tracer species, whose atmospheric lifetime is so long that the resident concentrations are predominantly determined by mixing and transport processes. Examples are CO2 and SF6, where the relative concentrations of both compounds are slowly increasing (with CO2 increases originating from the burning of fossil fuels, annual variability in the ‘sink’ terms and deforestation, and SF6 increases coming from power generation). These trace species are different in that CO2 has a seasonal cycle imposed due to biogenic activity, which predominantly occurs in the Northern Hemisphere. Dr Harald Bönisch (now at the J.W.Goethe University in Frankfurt) and Dr Bram Bregman (now climate co-ordinator at KNMI), in collaboration with the University of Leeds, have used measurements of these trace gases taken during the SPURT campaign between 2001-2003 to quantify the ability of st-TM5, TOMCAT and SLIMCAT to capture transport across the UTLS in the mid-latitudes. They find that although the models can capture the seasonal cycle, they are a few months out of phase in the LS as a result of inadequacies in the description of the tropospheric mixing. Figure 1 below shows the differences in [CO2] simulated using st-TM5 for selected months in 2001, where the seasonal cycle can be clearly seen. Figure 2 shows comparisons of interpolated tracer fields from all three CTM’s with SPURT measurements for four selected days in 2001/2002. It shows that for different seasons that ability of each model to capture the observed variability changes significantly.