Although the diffusion capacity of cigarette papers is routinely measured and instruments for this measurement have been available for several years, the repeatability and reproducibility of diffusion capacity measurements are not yet fully satisfactory. One reason for the high variation in diffusion capacity data are deviations in temperature and pressure from standardised conditions, especially within the measurement chamber of the instrument. Deviations in temperature and pressure will have an influence on the gas flow rates, the measured CO₂ concentration and the diffusion processes inside the measurement chamber. An evaluation of these effects on a theoretical basis shows that they cause relative errors in the diffusion capacity of several percent for temperature deviations of about 5 degrees or pressure deviations of about 5 kPa. Usually, errors are larger for cigarette papers with high diffusion capacity. The individual effects may add up to a relative error of more than 20%. As a mathematical model of diffusion and convection processes in the measurement chamber is often used in instruments to calculate the diffusion capacity from the measured CO₂ concentration, these errors can be reduced by modifying the existing model and including temperature and pressure effects. The dependence of model parameters on temperature and pressure was derived from data available in the literature. The modified model is evaluated and it is shown that the errors can be substantially reduced. An important conclusion is also that a meaningful statement of diffusion capacity values should include temperature and pressure during measurement.