PURPOSE: To determine if addition of the ulcer-coating polysaccharide sucralfate could improve symptomatic relief of radiation mucositis over a popular combination of antacid, diphenhydramine, and viscous lidocaine alone. METHODS AND MATERIALS: A double-blind study was conducted in which nurses and pharmacists coded patient groups and distributed medication in a manner blinded to both the patients and physicians. Eligible patients receiving radiation to the head and neck and/or chest sites that included the esophagus were randomized to a standard combination of antacid, diphenhydramine, and viscous lidocaine vs. the same solution plus sucralfate. Eligible patients were those receiving >40 Gy at 1.8 Gy/fraction, one fraction/day, five fractions/week. Participating patients were stratified between chest, small field head and neck, and large field head and neck. The patients subjective evaluation of throat soreness and relief with medication was elicited as well as physician observations and smears for Candidiasis screening. Medication was prescribed when the patient became symptomatic and concomitant use of other locally effective nonstudy agents was not allowed. The ability to eat various consistency of foods was graded 0-5, with 5 indicating no compromise of ability to ingest a food compared to baseline. Statistical analysis included mean + SD for food and soreness scores, paired t-test, and two-way analyses of variance to evaluate effects of site and treatment group on the changes in scores. RESULTS: Over 2 years, 111 patients were entered. Because some withdrew and others did not require medication, results are presented for evaluable patients in each category. Mild adverse effects from the medication solution (usually mouth discomfort) were reported by <10% of patients in each treatment group among 106 patients evaluable for toxicity. There was a comparable incidence of mild-moderate mucositis for the two treatment groups. Severe mucositis was noted in two patients of the standard medication group and none among patients receiving sucralfate. The groups were comparable for indicators including degree of soreness, dietary changes, and objective measures or mucosal irritation or infection. For the head and neck patients there was a significant worsening in soreness and in the ability to eat in both treatment groups, whereas patients treated to the chest had less nutritional change. Multivariate analysis including control for confounding factors did not detect treatment effects for any of the response measures. No patient had occult fungal infection detected by oral pathology evaluation of routine mucosal scrapings. CONCLUSIONS: A trend (that was not statistically significant) of less severe radiation mucositis was noted for patients receiving sucralfate in addition to the combination of viscous lidocaine, diphenhydramine, and antacid for nonulcerative radiation mucositis.