BACKGROUND: In South Africa, there are no national guidelines for the conduct or quality assessment of colonoscopy, the gold standard for investigation and diagnosis of bowel pathology. OBJECTIVES: To describe the clinical profile of patients and evaluate the practice of colonoscopy using procedural quality indicators at the Wits Donald Gordon Medical Centre (WDGMC) outpatient endoscopy unit (OEU). METHODS: We conducted a prospective, clinical practice audit of colonoscopies performed on adults (≥18 years of age). A total of 1 643 patients were included in the study and variables that were collected enabled the assessment of adequacy of bowel preparation, length of withdrawal time and calculation of caecal intubation rate (CIR), polyp detection rate (PDR) and adenoma detection rate (ADR). We stratified PDR and ADR by sex, age, population group, withdrawal time and bowel preparation. CIR, PDR and ADR estimates were compared between patient groups by the χ2 test; Fisher's exact test was used for 2 × 2 tables. A p-value <0.05 was used. Benchmark recommendations by the American Society for Gastrointestinal Endoscopy (ASGE)/American College of Gastroenterology (ACG) Task Force on Colorectal Cancer (CRC) were used in this audit to assess individual endoscopist performance and that of the endoscopy unit as a whole. RESULTS: The mean age of patients was 55.7 (standard deviation (SD) 14.4; range 18 - 91) years, ~60% were female, and the majority (75.5%) were white. Of the outpatients, 77.6% had adequate bowel preparation (ASGE/ACG benchmark ≥85%). The CIR was 97.0% overall, and screening colonoscopy was 96.3% (ASGE/ACG benchmark ≥90% overall and ≥95% for screening colonoscopies). The median withdrawal time for negative-result screening colonoscopies was 5.7 minutes (interquartile range (IQR) 4.2 - 9.3; range 1.1 - 20.6) (ASGE/ACG benchmark ≥ 6minutes), and PDR and ADR were 27.6% and 15.6%, respectively (ASGE/ACG benchmark ADR ≥25%). We demonstrated a 23.7% increase in PDR and 14.1% increase in ADR between scopes that had mean withdrawal times of ≥6 minutes and <6 minutes, respectively. Although the number of black Africans in the study was relatively small, our results showed that they have similar ADRs and PDRs to the white population group, contradicting popular belief. CONCLUSIONS: The WDGMC OEU performed reasonably well against the international guidelines, despite some inadequacy in bowel preparation and lower than recommended median withdrawal times on negative-result colonoscopy. Annual auditing of clinical practice and availability of these data in the public domain will become standard of care, making this audit a baseline for longitudinal observation, assessing the impact of interventions, and contributing to the development of local guidelines.