A formulation of 2% cromolyn sodium (CS) ophthalmic solution without the preservative, 2-phenylethanol, was compared with placebo in 58 patients with seasonal allergic conjunctivitis. Selection was based on history and positive skin tests. Neither immunotherapy nor use of antihistamines was allowed. This study was double-blinded and stratified by RAST scores to assure comparable groups. Either CS or placebo was used six times daily. Patients were observed weekly for 5 weeks during the peak of the fall weed-pollen exposure. Nasal symptoms were treated as required with beclomethasone nasal spray, and uncontrolled ocular symptoms were treated with boric acid and ephedrine solution. Nasal and ocular symptoms were recorded. There was a significant suppression of eye symptoms in the group receiving CS ophthalmic solution (p less than 0.02) during weeks 2, 4, and 5. There was a trend for nasal symptoms and the requirement for nasal beclomethasone to be less in patients receiving CS.