The differential diagnosis between untreated ulcerative colitis (UC) and Crohn’s disease (CD) in resected specimens is usually not difficult. UC is a diffuse, continuous mucosal disease that begins in the rectum and spreads variably to the proximal colon. The disease is worse distally. The ileum is not involved. Granulomas, fissures, and transmural lymphoid aggregates are not present. CD shows segmental distribution with variable disease severity (worse proximal) and variable rectal involvement. The ileum is involved in up to 40% of the patients. Granulomas, unrelated to mucin or foreign bodies, and fissures may be present. Fibromuscular obliteration and nerve fiber hyperplasia in the submucosa and transmural lymphoid aggregates are common. In specimens from patients operated on for severe (fulminant) colitis, a definite diagnosis can be difficult because of overlapping features. These cases are classified as “colitis of uncertain type — etiology” (indeterminate colitis).1 The differential diagnosis between UC and CD is important in these patients because an ileal pouch anal anastomosis is often recommended. In CD this type of operation may be less appropriate. Transmural lymphoid aggregates (specificity: 90%; positive predictive value: 62%), granulomas in the subserosa, and submucosal nerve fiber hyperplasia with ganglionitis all favor CD.2 The endoscopic biopsy diagnosis of UC and CD remains a problem. Many features useful for the diagnosis when surgical resection specimens are available are not present or are nonspecific when only the mucosa and superficial submucosa are available. Additionally, medical treatment changes the typical disease patterns. The two diseases share features and neither disease has a pathognomonic finding present in every case of the one and absent in every case of the other. The relative absence of discriminating features can partly be solved by analysis of multiple endoscopic biopsies from different areas of the colon and ileum.3 This allows conclusions about the distribution of the disease, the focal nature of inflammation in the same area, and the presence of areas of sparing. Discontinuous disease is a major indicator that the underlying condition is unlikely UC.4 Stepwise biopsies also provide information about the variability of the disease (UC is worse distally). Ileal involvement favors CD. Ileitis in UC is present only when active total disease involves the ileocecal valve in continuity with distal disease. Full colonoscopic biopsy series provide a correct diagnosis in 64% of cases for CD and 74% for UC, increasing to 96% with the addition of clinical information. (5) The clinical presentation can be atypical and a precise diagnosis remains difficult in early-onset disease (2–6 weeks after the first symptoms), in children, especially young children below the age of 12, and in patients with associated liver disease (hepatitis, primary sclerosing cholangitis).1 These patients can be classified temporarily as colitis, type unclassified (IBDU). The microscopic criteria applicable to endoscopic biopsy specimens for the distinction of inflammatory bowel disease (IBD) from other forms of colitis and of CD from UC include a combination of architectural and inflammatory features and other lesions. Architectural abnormalities are crypt distortion (nonparallel, variable diameter, cystic), crypt branching, crypt shortening, decreased crypt density, and irregular mucosal surface. Inflammatory features are transmucosal increase in lamina propria cellularity (basal plasmacytosis accumulation of lymphocytes and plasma cells between the bases of crypt and the muscularis mucosae or close to the muscularis mucosae) and neutrophils. Neutrophils are mainly an indicator of disease activity. Other lesions include a change in epithelial mucin content, abnormalities of the muscularis mucosae, and metaplasia (pseudopyloric metaplasia in the ileum and Paneth cell metaplasia in the colon). Abnormal architecture and basal plasmacytosis distinguish between chronic idiopathic IBD (present) and other causes of colitis (absent). In this regard, diverticular disease associated colitis is a differential diagnostic problem.6,7 These features may not be present at the initial onset of the disease. Plasmacytosis increased from 38% in biopsies obtained 1–15 days following onset of symptoms to 89% in biopsies obtained after 121–300 days. Architectural abnormalities increased from 0%–78%.8 Metaplasia occurs mainly in longstanding disease. After recovery with or without treatment, basal plasmacytosis and architectural abnormalities may decrease. Repeat endoscopy with biopsy and comparative study From the Department of Pathobiology, University Hospital, Leuven, Belgium. Copyright © 2008 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1002/ibd.20598 Published online in Wiley InterScience (www.interscience.wiley.com).