Adsorption in microporous materials is important in applications such as mixture separations, gas storage and catalysis. Conversely, physical adsorption is used in the characterisation of porous materials. Some of the practical difficulties facing experimentalists are discussed and the advantages of using simulation as an interpretative tool advanced. The remainder of the review concentrates on equilibrium adsorption. After a brief outline of the relevant methodology, several specific examples are addressed. In the first of these, high (i.e. ambient) temperature adsorption of nitrogen in a well characterised activated carbon fibre is discussed. It is demonstrated that high temperature adsorption offers advantages in characterisation over the conventional low-temperature method, but that care is needed in evaluating the total adsorption from the surface excess. A second example is the adsorption of ethane–methane mixtures in graphite pores. The selectivity S turns out to be a highly sensitive property in relation to the type of model adopted. In simulations where a two-centre ethane model is employed, entropic factors are very important. This general conclusion is supported by simulations with artificially elongated ethane and by studies of ethane–propane mixtures. In the final example, the low-temperature adsorption of argon in silicalite is the focus of attention. A new argon–silicalite potential function has been developed, and compared with the frequently used potential originated by Kiselev and co-workers. The new potential model is consistently superior in predicting zero coverage heats and Henry law constants. However, it turns out that neither potential, when used as input to a simulation, can successfully predict all of the isotherm transitions observed experimentally. The explanation proposed is that real adsorbents, unlike the rigid solids most frequently modelled in simulations, can undergo distortion in some systems when critical adsorbate loadings are reached.