Abstract Rationale Low dose topical corticosteroids (CS) have either no effect on growth velocity (GV) or transient effects without impact on final adult height. Although many growth studies have been conducted; the relationship between CS exposure and GV, the no-effect exposure level and the relative sensitivity of GV versus cortisol suppression as markers of CS exposure, are poorly defined. Methods The data from 32 published stadiometry and knemometry studies were collated and CS exposure estimated in cortisol equivalents (CE, pg.h/mL). This was compared to the annualized ▵GV data (cm/y). Results ▵GV and CE were highly correlated and described by a non-linear sigmoid Emax model. The model predicted rank order potential to reduce GV, expressed as multiples of the pediatric dose (μg/d) were: oral prednisolone (5000μg/d) 0.14 > inhaled beclomethasone (400μg/d) 0.55 > inhaled budesonide (400μg/d) 0.68 > intranasal triamcinolone (220μg/d) 0.77 > inhaled triamcinolone (400μg/d) 0.78 > intranasal beclomethasone (336μg/d) 0.89 > inhaled mometasone (200μg/d) 2.0 > intranasal budesonide (128μg/d) 2.6 > inhaled fluticasone (200μg/d) 2.8 > intranasal mometasone (100μg/d) 120 > intranasal fluticasone (100μg/d) 150. Values >1 predict no significant impact on GV. The model also predicted that a detectable reduction in plasma cortisol should be apparent when ▵GV falls below −0.8cm/y, an equivalence endpoint used in growth studies. Conclusions Since ▵GV was non-linearly related to CS exposure, no-effect dose regimens are possible for CS with low systemic exposure. It also appears unlikely that growth inhibition should occur in the absence of detectable reductions in cortisol.