Background NSAIDs are commonly used analgesic agents in the orthopaedic trauma setting. Evidence-based guidelines recommend that patients with one or more risk factors for NSAID-associated gastrointestinal (GI) ulcer complications should be prescribed gastroprotective agents to minimise the risk of serious ulcer complications, including gastrointestinal haemorrhage. The purpose of the present audit was to evaluate and improve the adherence to these guidelines in new-NSAID users in a trauma unit at a district general hospital. Methods A retrospective observational cohort study was conducted over an 18-week period to assess pre-intervention practice. Subsequently, an awareness programme, including prescriber and pharmacist education and the use of reminder posters, was implemented. Following this, data were collected prospectively over 9 weeks to assess any change in performance. Assessment involved review of case-notes and prescription charts of all adults (aged ≥18 years) who were commenced on regular NSAIDs on or during admission to the Trauma Unit. Patients were risk-stratified according to the number of risk factors, which were defined as age ≥65 years, major comorbidity, oral steroids, anticoagulation, history of upper gastrointestinal ulceration or bleeding and prescription above the normal recommended dose of NSAIDs. The American College of Rheumatology guidelines recommend the use of gastroprotective agents when one or more risk factors was present. Prescription of gastroprotective drugs was recorded to measure adherence to evidence-based guidelines. Results A total of 644 patients were reviewed over the study period, 451 pre-intervention and 193 post-intervention. 100 patients fulfilled the inclusion criteria pre-intervention and 49 post-intervention. Before intervention, the proportion of high-risk NSAID-receivers co-prescribed gastroprotection was low at 25.3%, although the likelihood of adherence improved with the number of risk factors; overall adherence rate improved significantly following intervention at 73.1% ( χ 2 = 18.8, p < 0.001). Furthermore, a smaller proportion of NSAID-receivers fell into the high-risk category from 75% to 56.5% ( χ 2 = 7.25, p < 0.05). Conclusions (1) The majority of trauma admissions are at high risk for developing gastrointestinal haemorrhage. (2) Initial adherence to national guidelines for safe prescription of NSAIDs in our trauma unit was poor (25.3%) but improved significantly (73.1%) following an awareness programme which included education of prescribers and pharmacists. (3) A lower proportion of NSAID-receivers had multiple risk factors following our awareness programme. (4) Awareness of gastroprotection guidelines must be raised in trauma units to prevent undertreatment and hence minimise the risk of GI haemorrhage.