A double-blind study revealed that: (1) a commercially available solution of 2% tripelennamine and 0.5% chlorobutanol gave a significant degree of pharyngeal anesthesia as measured by gag reflex; (2) a placebo group without pharyngeal anesthesia was endoscoped without significant difficulty; (3) neither the presence or absence of gag reflex nor the use of topical anesthetic agent made a significant difference in the difficulty of esophagogastroscopy, except that (4) those patients who appeared to the endoscopist to be likely to experience difficulty in the procedure had significantly less difficulty after gargle with the anesthetic agent.