Conventional models of the immune response are based on distinguishing self and non-self. However, we consider that the more recently proposed 'danger' model may be an illuminating alternative for studying allergic contact dermatitis. In this model, an antigenic signal on its own would tend to produce tolerance. In contrast, in the presence of a 'danger' signal, which, in the case of allergic contact dermatitis, we suggest is usually cutaneous irritancy, the immune system would become activated, leading first to the induction of sensitization and then subsequently to the elicitation of a contact hypersensitivity response. In most cases, both the antigenic signal and irritant signal will come from the hapten, although, e.g., in an occupational setting, traumiterative dermatitis would be the source of the 'danger' signal. Typically, the irritant signal tends to be more concentration-dependent and thus is the overriding factor in the determination of the effective sensitizing and eliciting concentrations of the hapten. A further prediction of this hypothesis is that successful experiments demonstrating low-dose tolerance with contact allergens may be explained by the loss of the irritant effect at lower dilutions, whilst an antigenic stimulus remains present.