The effect of a corticosteroid drug on type I reactions in the nose was studied in eighteen hay fever patients. They were treated with a daily dosage of 800 microng beclomethasone dipropionate intranasally for I week, in a double-blind, placebo-controlled cross-over trial. Allergen challenge on the last day of each treatment period showed that steroid treatment diminished sneezing and rhinorrhoea, but the effect was not statistically significant (P = 0-10). Immediate nasal blockage was unaffected. There is a marked discrepancy between the minor protection offered by beclomethasone dipropionate in type I allergy, as suggested in this study and the marked efficacy of the drug in the treatment of hay fever. It is suggested that two factors may account for this discrepancy. First, the significance of "delayed" reactions for hay fever symptoms. Second, the difference between a single allergen provocation and continuous exposure. It is suggested that a slight steroid inhibition of immediate reactions following allergen provocation is potentiated during natural exposure due to inhibition of the priming of the end-organ.