A double-blind comparison of betamethasone valerate and sodium cromoglycate both given by the nasal route was carried out in forty patients with seasonal rhinitis caused by grass pollen. All patients kept daily symptom score cards, and half of them measured both oral and nasal peak expiratory flow rates twice daily. Adrenal function was monitored in thirty-one patients and found to be normal throughout. Sixteen of those patients receiving the steroid aerosol recorded success and two failure of treatment. By contrast, of those receiving sodium cromoglycate there were only two treatment successes and twelve failures. The total symptom score recorded in the group receiving betamethasone valerate was about half that recorded by the sodium cromoglycate group (P less than 0.01). No difference was observed between the two treatments in respect of nasal peak flow rate; specific IgE blood levels and weal sizes following prick tests were not significantly altered throughout the period of the trial, although total IgE was significantly increased. These results suggest that nasal betamethasone valerate offers patients with allergic rhinitis marked symptomatic benefit without the disadvantages previously associated with steroids.