In this double-masked, double-dummy, randomized, single crossover study, we compared single doses of a fast-dissolving wafer formulation of famotidine with a conventional tablet formulation of ranitidine in patients with gastroesophageal reflux disease (GERD). Patient preference time until symptomatic relief, and predictive characteristics of early responders were assessed. Eligible patients had a clinical diagnosis of GERD and symptoms of GERD of sufficient severity to require relief. The study treatment was one dose of famotidine (20-mg wafer) and one dose of ranitidine (150-mg tablet), which were given in a randomized order and taken as needed. The patients were instructed to measure the symptomatic effects on a seven-point categorical scale (1 = worse to 7 = free of symptoms) at 15, 30, 45, 60, 120, and 180 minutes. After the clinical phase of the trial, the patients indicated their global assessment of efficacy and their preference for the wafer or the tablet. Of the 829 patients who completed the study, significantly more preferred the wafer to the tablet. While there was no significant difference in the global assessment of efficacy, the famotidine wafer provided significantly better relief than the ranitidine tablet during the first hour after dosing. However, at 120 and 180 minutes, the degree of relief was similar for the two drugs. The time until a clinically significant effect was also similar for the two drugs, and approximately one half of the patients experienced such improvement within 3 hours. Multivariate analyses disclosed no predictive characteristics of early symptomatic effect.