Experimental results of the performance of a pneumatic suspension fitted in a single wheel station test rig are presented. Results for both a passive and slow–active version of the suspension are compared; the slow–active version involves an additional pneumatic actuator which has limited frequency response capabilities. The slow–active system is shown to provide worthwhile improvements in ride performance compared with the passive system. A theoretical model of the slow–active suspension is developed and predicted results are compared with experimental measurements on the rig. The mathematical model produces optimistic results for both passive and slow–active suspensions and the practical limitations of the experimental hardware are discussed.