AIM: To assess the performance of a histopathology department in diagnosing inflammatory bowel diseases with comparison of reports from other centres. STUDY POPULATION: 1067 sets of endoscopic biopsies received in the department of histopathology, Royal Hallamshire Hospital, 1990-1995. METHODS: The histopathological diagnosis of non-neoplastic endoscopic colorectal biopsies was audited using data from histopathology reports. The biopsy diagnosis by the initial reporting pathologist and final diagnosis after additional investigations (endoscopy, radiology, microbiology) or surgery were used to derive sensitivity, specificity, and positive predictive values for categories of disease. RESULTS: Diagnosis was validated for 1067 biopsy sets (43% of those initially assessed). For all biopsies (with or without active inflammation) reports highly suggestive or suggestive of Crohn's disease had a sensitivity of 50%; for ulcerative colitis the comparable figure was 62%. Sensitivity was the same for both diagnoses (74%) in those biopsies with active inflammation. Positive predictive values for highly suggestive diagnoses of ulcerative colitis or Crohn's disease were 100%. In all biopsies the specificity of a histopathological diagnosis of normality was 96%. CONCLUSIONS: These results compare favourably with the other published audits and present an achievable level of performance for non-specialist hospitals with non-specialist histopathology services.