AIM To examine the accuracy of data collected on pressure ulcer incidence. A system for monitoring incidence has been in place since 1991 at a medium-sized hospital trust. The data collected are used to calculate resource requirements for specialist mattresses, to audit compliance with pressure ulcer guidelines and to identify staff training needs. METHOD A convenience sample comprising 211 inpatients from the medical, surgical, rehabilitation and orthopaedic wards was assessed by the researcher over a 14-day period. Ward staff continued their usual routine of assessment and treatment. At the end of the study period, the researcher-generated data and ward staff-generated data were compared. RESULTS A total of 343 assessments were carried out on 211 patients; 152 patients were found to have, or be at risk of having, pressure ulcers. The main finding after comparing both sets of data was that a significant number of patients at risk of, or with, pressure ulcers were not reported as such by ward staff (n = 86, 57 per cent; p < 0.0001). Twelve (14 per cent) of those not reported had pressure sores and were being cared for on specialist mattresses. Non-reporting was high from all wards. When reports were made of patients at risk of, or with, pressure ulcers, the reports were mostly accurate and complete. CONCLUSION The data collection form is user-friendly and the assessment skills of nursing staff at the trust are adequate. However, the high incidence of under-reporting indicates that the current system may not be adequate for generating accurate data.