The identification of chemicals possessing the intrinsic ability to cause sensitization via skin contact or inhalation, commonly referred to as skin and respiratory sensitizers, is a key endpoint in regulatory toxicology. Predictive assays for this purpose exist only for skin sensitizers, but, for both types of sensitizer, human evidence can be used to determine whether a substance should be classified. Furthermore, the use of human evidence for subcategorization according to sensitization potency is also accommodated within the regulations. Normally, this is based on the prevalence of sensitization in relation to the degree of exposure in the context of the size of the population exposed. However, the regulations also indicate that the severity of (allergic) reactions may be taken into account. In this article, we consider whether this is appropriate and whether there is evidence that reaction severity can inform decisions on classification and/or potency categorization. The conclusion drawn is that the severity of an allergic reaction does not correlate with, or serve as an indicator of, the sensitizing potency of a chemical. In reality, it reflects the overall extent of sensitization that an individual has acquired, in concert with the concentration of the causative allergen to which they have been exposed.