OBJECTIVES: To determine whether giving patients a resource use log (RUL) at hospital discharge reduces recall bias in a follow-up resource use questionnaire (RUQ).  METHODS: Within the APEX randomised controlled trials (RCTs), 86 patients undergoing joint replacement were randomised to receive or not receive an RUL at hospital discharge (The RUL trial). A postal RUQ was then administered to all participants at 3-months after surgery.   Resource use data (visits to GPs, GP home visits and telephone calls; GP practice nurse visits and telephone calls; and prescribed medication) in relation to the patient’s joint replacement were extracted from GP records from hospital discharge until completion of the 3-month RUQ by a blinded researcher. Data from both sources were coded into use of resource and number of contacts. For each resource use category, descriptive statistics were calculated by data source and RUL trial arm. Kappa statistics and Concordance Correlation Coefficients (CCC) were calculated as appropriate.  RESULTS: GPs were contacted for 67/86 patients originally randomised to receive or not receive an RUL (one had surgery delayed, 3 died, 5 withdrew, 6 had GP practices outside of area, 4 did not complete the 3-month RUQ). Information was then extracted for 66/67 patients. There was evidence of improved recall in favour of the RUL arm in relation to visiting a GP (Kappa=0.5312 vs.-0.0161). There was some slight evidence in favour of the non-RUL arm with regards to having a GP home visit (Kappa=0.335 vs. -0.0937). The RUL arm showed more agreement than the non-RUL arm between data sources in terms of number of: visits to GPs (CCC=0.581 vs. -0.013); GP telephone calls (CCC=0.564 vs. 0.173) and prescriptions (CCC=0.418 vs. -0.13). CONCLUSIONS: Although based on small numbers, our study found some evidence that provision of an RUL reduces recall bias in relation to visits to GPs.