BACKGROUND: We have previously demonstrated the clinical efficacy of montelukast in a randomized double-blind controlled cross-over trial in patients with dyspepsia in association with duodenal eosinophilia. The mechanism of this clinical response is unknown but could involve a decrease in eosinophil density or activation. METHODS: Twenty-four dyspeptic patients 8-17 years of age underwent initial blood sampling and endoscopy with biopsy. Eighteen of these patients had elevated duodenal eosinophil density and underwent repeat blood sampling and endoscopy following 21 days of therapy with montelukast (10 mg/day). The following were determined: global clinical response on a 5-point Lickert-type scale, eosinophil density utilizing H & E staining, eosinophil activation determined by degranulation indices on electron microscopy, and serum cytokine concentrations. On day 21, pharmacokinetics and duodenal mucosal drug concentrations were determined. RESULTS: Eighty-three percent of the patients had a positive clinical response to montelukast with regard to relief of pain with 50% having a complete or nearly complete clinical response. The response was unrelated to systemic drug exposure or to mucosal drug concentration. Other than a mild decrease in eosinophil density in the second portion of the duodenum, there were no significant changes in eosinophil density, eosinophil activation, or serum cytokine concentrations following treatment with montelukast. Pre-treatment TNF-alpha concentration was negatively correlated with clinical response. CONCLUSION: The short-term clinical response to montelukast does not appear to result from changes in eosinophil density or activation. Whether the effect is mediated through specific mediators or non-inflammatory cells such as enteric nerves remains to be determined. TRIAL REGISTRATION: ClinicalTrials.gov; NCT00148603.