Some drug induced immune mediated side effects appear rapidly (6-<48hrs) and are hard to reconcile with the generation of a new immune response to the drug. By extending the previously presented p-i concept (pharmacological interaction with immune receptors), I propose that drugs might stimulate memory T-cells via their T-cell receptors, which happen to react not only with a peptide antigen, but also with a (chemically inert) drug. In this model, the generation of a new, drug specific immune response is actually bypassed: the innate immune system must not be stimulated, as no naive T-cells are stimulated. Only previously activated memory T-cells are re-activated, as they have a lower threshold of reactivity to a T-cell receptor transmitted signal, whereby certain cofactors like an ongoing immune response to a virus might facilitate reactivity to the drug. This concept has major implications for preclinical testing of the allergenic potential of a drug, which normally is measured by the ability of the drug to cause a new immune response.