The demographic characteristics of patients used in clinical trials (such as the severity of airway obstruction) can significantly influence the results of dose-response studies, emphasising the need to evaluate effects on the steep part of the dose-response curve. Differences in inhaler devices can also influence study outcomes, as for inhaled drugs both airway efficacy and adverse effect profiles are primarily determined by lung deposition and hence bioavailability. Dose-response studies with short- and long-acting beta 2-agonists show an excellent therapeutic ratio at conventional doses used in everyday clinical practice (i.e. 2 to 4 puffs). Dose-related systemic effects of beta 2-agonist occur at higher doses, for salbutamol (albuterol) > 500 micrograms. Fenoterol is a beta 2-agonists with higher intrinsic activity than salbutamol and produces greater systemic effects at higher than conventional doses on a microgram equivalent basis, although even at 4000 micrograms such differences are unlikely to be clinically relevant. No differences between fenoterol and salbutamol have been shown in terms of bronchodilator potency on a microgram equivalent basis. The long-acting beta 2-agonist salmeterol, as a partial agonist, has the potential to attenuate the acute bronchodilator response to a higher activity beta 2-agonist such as salbutamol or fenoterol, although there is no evidence to date on whether this is relevant in the setting of acute asthma. When comparing inhaled corticosteroids, attention should be focused on their respective risk-benefit ratios for antiasthmatic versus systemic activity. In terms of detecting systemic activity, it is important to use sensitive measures, such as urinary cortisol excretion, rather than insensitive parameters, such as a single morning plasma cortisol measurement between 0800h and 1000h. For fluticasone, a greater in vitro potency results in only marginal differences in antiasthmatic efficacy, particularly on the flatter part of the dose-response curve above 1000 micrograms/day in adults and 400 micrograms/day in children. However, the same enhanced potency translates directly into commensurate differences in systemic adverse effects on the steep part of the systemic dose-response curve above 1000 micrograms/day in adults and 400 micrograms/day in children, respectively. Furthermore, with repeated twice-daily administration, a longer elimination half-life and prolonged systemic tissue retention due to enhanced lipophilicity will result in greater systemic activity observed at steady-state in long term administration studies. This dissociation of airway and systemic dose-response curves results in a J-shaped curve for benefit: risk ratio, with a watershed area above 1000 microgram/day in adults. This fall in the benefit: risk ratio is likely to be greater for fluticasone than for budesonide or beclomethasone. Further studies are needed to clearly define the dose-response relationships of higher potency steroids such as fluticasone, particularly on the steep part of the curve (for clinical efficacy), using the appropriate back-titration design along with sensitive measures of antiasthmatic and systemic activity.