A method of assessing the allergic status of a subject to a specific allergen, comprises the steps of determining a subject-specific numeric value for a plurality of variables comprising age, sex, SPT, IgE, (sIgE or tIgE minus sIgE) and allergic symptoms, applying an allergen-specific weighting to the numeric value for each variable, and correlating the weighted numeric values with allergic status for the subject. The weighted numeric values may be correlated with allergenicity status using a linear regression model P = ez/(1= ez) in which: P = predicted allergenicity status for the subject; e = mathematical constant (2.71828....); and z = linear combination of the weighted numeric values for the variables.