The evidence for hypothalamic-pituitary-adrenal axis (HPA) suppression by inhaled corticosteroids (ICS) was found to be conflicting. Reviewers have not distinguished between gold standard and basal adrenal function tests. The utility of the latter is limited by physiological and pathological variability as well as by methodological concerns. The risk of HPA suppression in asthmatic children and adolescents treated with ICS, as determined by gold standard adrenal function tests, needs to be established. A systematic review of the literature from January 1973 to July 2005 was performed. The Medline and Cochrane databases were searched, the reference lists of retrieved articles were inspected and pharmaceutical companies were approached. Randomized-controlled trials, cohort and case-control studies designed to detect HPA suppression caused by ICS, diagnosed by the insulin tolerance test (ITT) or the metyrapone test, performed on asthmatics of all ages not on oral steroids, were included and assessed for methodological quality. Of the 22 identified studies only four met the criteria for inclusion. All of these were published before 1988 and only one was methodologically sound. The cohort study showed that the baseline risk for HPA suppression is 0% while the absolute risk is 100% in asthmatic children treated with a beclomethasone dipropionate metered dose inhaler at a dose of 250-600 mug/m(2)/day for 6-42 months. As suggested by other observations these results could be generalized to other ICS. They may be of clinical significance especially if children are subjected to stress. Further research is needed to establish the cumulative dose for all ICS at which HPA suppression will be precipitated. Guidelines for future trials are suggested.