Data from experiments, in which colonies of a hydroid, Laomedea flexuosa, were exposed to a range of Cu2+ concentrations and a marine yeast, Rhodotorula rubra, was exposed to a range of Cd2+ concentrations, not only exhibit hormesis, but also suggest how its occurrence in growth experiments might be explained. When growth data are considered as normalized specific rates against a time base, their oscillatory form indicates the output of a growth regulatory mechanism whose behaviour can be used to interpret the typical concentration-response curve exhibiting hormesis. Advantages may be conferred upon organisms whose growth control mechanisms overcorrect in response to low levels of inhibitory loading by toxic agents (stimulus), while at higher concentrations it is the overloading of such control mechanisms that results in the threshold in concentration-response curves (inhibition). It is suggested that if different examples of hormesis share a common explanation, it lies in the behaviour of homeostatic and homeorhetic feedback mechanisms, which respond to perturbation non-specifically and may overcorrect for adaptive reasons to low levels of inhibitory challenge.