Oral mucositis (OM) has substantial negative clinical, quality-of-life, and economic consequences for patients with haematologic malignancies who require myeloablative chemotherapy or radiotherapy. Uniform training in OM assessment is infrequent in clinical practice, so the true incidence and duration of OM are unknown. Nurses and physicians from the European Group for Blood and Marrow Transplantation recently undertook an audit of 214 patients (197 evaluable patients) treated at 25 centres, the Prospective Oral Mucositis Audit (POMA), to determine the incidence, severity, and duration of OM. To standardise the assessment of OM severity, the World Health Organization (WHO) Oral Toxicity Scale was used across centres. This article focuses on the quality control analyses that were conducted to ensure that OM was accurately assessed across all 25 centres. Twenty-two trainers, who received comprehensive training about POMA study design, pathobiology of OM, and endpoint assessment, educated staff at the 25 transplantation centres about OM assessment. The trained staff collected data by completing daily worksheets for each patient. Three quality control analyses, of 82, 1949, and 4111 worksheets respectively, showed a nurse assessment accuracy rate of 74%, 90%, and 90%. The most common errors were in assigning WHO grade 0 or 1. This analysis shows that training of nursing staff had a positive effect on assessment of OM severity, which should ultimately lead to improvement in the quality of supportive care.