Using Hospital Episode Statistics (HES) data from England for the period 1996-2006, we performed a descriptive study to compare records of Clostridium difficile for inpatients aged >or=65 years and for all patients following any of four types of orthopaedic procedures. Results showed that infection rates for C. difficile increased whereas rates for orthopaedic surgical site infections (SSIs) decreased. Both types of infection were more common in older female patients and in patients with greater comorbidity, but showed little difference in rates between areas with varying deprivation scores. For 2004 and 2005, we compared the HES data with mandatory reporting data from the Health Protection Agency (HPA). This showed recording of C. difficile infection to be higher from HPA data than from HES data. In contrast, compared with HPA data for orthopaedic SSIs, there were many more SSIs and numbers of procedures recorded from HES data for all four orthopaedic procedures, although the infection rates themselves were broadly similar. These findings reflect the limitations of both methods used and we suggest that there is a case for using both sources of information, either independently or linked at an individual level in order to obtain a more complete picture of these important healthcare-associated infections. If better coding could be encouraged or made mandatory within HES data, then the current dual system of recording might be unnecessary for effective surveillance of orthopaedic SSIs.