To evaluate the clinical significance of colonic eosinophilia, we conducted a retrospective study of all children older than 1 year evaluated at Children's Hospital of Pittsburgh from January 1999 to June 2001 with a description of colonic eosinophilia in the pathology report. Medical records were reviewed. Diagnoses were confirmed by contacting the patients. Histological slides (H&E) were reviewed by an investigator blind to the patients' data. Biopsies were grouped according to the site they were taken from and then screened at low power for areas of maximal eosinophilia for further quantitative analysis. Results of manual counts were validated by image analysis using Metaphor Image Analysis Software. Sixty-nine children with colonic eosinophilia were identified (36 male; mean age, 135.2 +/- 55.4 months). Their final diagnoses were inflammatory bowel disease in 32% (group A), irritable bowel syndrome in 33% (group B), food allergies in 10% (group C), and other diagnoses in 25% (group D). The maximal eosinophil count per crypt area was significantly (P < 0.05) higher in group A vs groups B, C, and D (34.8 +/- 17.1 vs 21.3 +/- 8.8, 25.4 +/- 16.7, and 24.2 +/- 9.7, respectively). The total cellularity of the lamina propria was considered high only in group A (P < 0.05 vs groups B and C). A mostly equal vertical distribution of eosinophils throughout the lamina propria was found significantly more frequently in group A vs groups C (P = 0.04) and D (P = 0.007). We conclude that children with inflammatory bowel disease have an equal distribution of eosinophils throughout the lamina propria, with intraepithelial and intracryptal eosinophils and with a higher overall total cellularity. In irritable bowel syndrome and patients with a variety of other diagnoses, including allergies, the distribution is mostly superficial, with a lower total cellularity.