Objectives:Functional dyspepsia (FD) is a highly prevalent gastrointestinal disorder with unknown etiology. We aimed to assess the association between dyspeptic symptoms and duodenal eosinophilia in children referred for upper endoscopy.Methods:In this retrospective cohort study, all histopathology reports of normal upper endoscopies performed at a single tertiary center between 2010 and 2014 were reviewed. FD was defined as epigastric pain or discomfort for more than 2 months with no response to acid suppressants. Controls were those with non‐erosive reflux disease, dysphagia or rumination syndrome. Duodenal biopsies were analyzed by pathologists blinded to indication. Intramucosal eosinophils were counted per mm2 on haematoxylin and eosin stained digital imaged sections using Aperio eSlide Manager. Demographics, clinical variables, family history and histologic findings were compared between cases and controls.Results:In total, 36 cases and 36 non‐matched controls were identified of which 56% and 53% were female (p = 0.81). Mean (± SD) age was higher in cases compared to controls (13.6 (± 3.1) vs. 10.5 (± 4.0) years; p = .001). Dyspeptic symptoms (epigastric pain 81% and/or upper abdominal discomfort 33%) were food‐related in 69% and nocturnal in 31% of cases. Self‐reported nausea (64% vs. 17%; p < .0001), lethargy (19% vs. 0%; p = .005) and a family history of functional gastrointestinal disorders (28% vs. 3%; p = .003) were more common in cases than controls. There was a higher but not significantly different rate of atopic history (39% vs. 25%; p = .21) and psychological co‐morbidity (53% vs. 39%; p = .24) in FD cases versus controls. Duodenal intraepithelial lymphocytes per 100 enterocytes were similar in cases and controls (median (IQR) 10 (8–13) vs. 12 (8–18); p = 0.12). Duodenal eosinophil counts per mm2 were significantly increased in cases compared to controls (151 (118–207) vs. 76 (60–106); p < 0.001).Conclusions:This study confirms recent reports of duodenal eosinophilia in FD. The high rate of atopic and psychological co‐morbidity in FD suggests multi‐factorial mechanisms and may explain why current therapeutic options aimed at symptom control are largely unsatisfactory. Duodenal eosinophilia in FD should be considered a therapeutic target.