Central venous catheter (CVC)-related infections (CRIs) are a key target for infection control in intensive care units (ICUs). The aim of this study was to describe temporal trends of CRI incidence in a network of volunteer ICUs in Northern France. During a 4 month surveillance period each year, all CVCs in place for more than 48h were prospectively followed until removal or patient discharge. Standard clinical and microbiological criteria were used to define colonization and CRI. The standardized incidence ratio (SIR) was estimated by dividing the number of observed CRIs by the number of expected CRIs, which was computed using a logistic regression model including risk factors for CRI. CRI incidence and SIR were fed back to ICUs as a benchmark at the end of each period. From 2001 to 2005, 135 ICUs participated for at least one surveillance period. Overall, 11 703 CVC in 9182 patients (122 495 CVC-days) were included. CRI incidence was 2.8 per 1000 CVC-days. Among 35 ICUs that participated for three or more consecutive periods, CRI incidence decreased significantly by 58.6%. SIR also decreased significantly from the first to the third surveillance period in these ICUs. These results suggest that surveillance programmes have a significant impact on CRI risk in ICUs and remain an important strategy for combating nosocomial infections in these settings.