OBJECTIVE: To examine the rates of co-prescribing of gastro-protective drugs (GPDs) and the frequency of gastrointestinal (GI) investigations associated with standard and slow-release (SR) formulations of nonsteroidal antiinflammatory drugs (NSAIDs). METHODS: Patients were identified from the MediPlus database, which contains the medical records of some 2 million patients throughout the United Kingdom. We selected all new NSAID users who received at least one prescription between 1 July 1997 and 31 December 1997. Patients were classified as starting SR or standard formulations by their first prescription and followed for 12 months. Events specified as of interest included GPDs prescribed and GI investigations. GPDs included prescriptions for all proton pump inhibitors, H2-antagonists and misoprostol. GI investigations included endoscopies and barium contrast studies. RESULTS: Taken overall, patients prescribed SR formulations of NSAIDs were significantly more likely to receive GPDs than those prescribed standard formulations, odds ratio (OR) 1.34, 95% confidence intervals (CI) 1.21-1.47. They also tended to be more likely to need GI investigations (OR 1.25, CI 0.91-1.71). Differences were not explained by prior predisposition to GI damage as judged by the previous need for GPDs, and were evident whether the NSAID prescription was the first, the second, or the third or later in the series (OR 1.40, CI 1.21-1.62; 1.33, 1.08-1.62; and 1.17, 1.00-1.38 respectively). CONCLUSION: This study found that patients prescribed SR formulations of NSAIDs were significantly more likely to receive GPDs, and may be more likely to need GI investigations than those receiving standard formulations. In addition, switching away from SR formulations was far more common than switching to them.