With the assumption that a provocation test gives the correct diagnosis, the possibility of predicting an allergy by means of the case history, a skin test, radioallergosorbent test (RAST) and combinations of these methods was evaluated for house dust, cow-, cat-, dog- and horse-dander, timothy-, marguerite-, dandelion- and birch-pollen allergens. All procedures were found to give a reliable diagnosis provided that a group with doubtful diagnosis was interposed between the negative and the positive ones. The doubtfoul group, requiring further diagnostic investigations, was smallest when all three methods were used in combination. It is suggested that, in addition to the case history, the first procedure in routine diagnosis with inhalant allergens should be a skin test. RAST should be used primarily for selected cases and as a complement to the case history and skin test, making provocation tests superfluous for the majority of patients.