SynopsisTo minimise consumer risk of allergic contact dermatitis, predictive testing may need to be carried out. An alternative approach examining physicochemical properties of substances, and from these trying to assess sensitization potential, formed the basis of a model described by Roberts and Williams [1]. The concept related electrophilic reactivity (k), the partition coefficient (P) and dose (D) in a term which described a relative alkylation index (RAI). This was shown to correlate well with experimental data for sultones and p‐nitrobenzyl halides: RAI =kD/(P±P2) (1)In recent studies of dose response relationships, the ‘overload effect’ predicted by Roberts and Williams is demonstrable. However, the simple relationship above omits two potentially important parameters, ‘intrinsic antigenicity’ of a chemical and activation in skin of an apparently unreactive chemical. Data on the sensitization potential of alkyl transfer agents shows that for small haptens, intrinsic antigenicity may depend in part on the nature of modifications to carrier protein tertiary structure. The ability of skin to metabolise 1,4‐substituted benzene derivativesin vivoappears even more complex than suggested [2], involving a spectrum of reactive intermediates characteristic of each molecule.These results suggest that the RAI formula needs to