The surveillance of healthcare-associated infections (HAIs) is a priority in intensive care, a high-risk setting due to patients’ critical condition and their exposure to invasive devices. Since 2004, the national REA-Raisin surveillance program, coordinated by the Network for Alert, Investigation, and Surveillance of HAI, has targeted infections associated with invasive devices in adult intensive care units for which a prevention strategy is essential: pneumonia (PNE), colonization, infection, or bacteremia (COL/ILC/BLC) associated with central venous catheters (CVC), and bacteremia (BAC). Since 2015, participating units have been collecting data continuously from January through December (compared to 6 months previously) for all patients hospitalized for more than 2 days. From January to December 2017, 199 departments included 68,581 patients (mean age: 64.2 years), hospitalized for an average of 10.4 days, of whom 70.0% were admitted to medical services, 17.8% to emergency surgery, and 12.2% to elective surgery; 8.1% of patients were trauma patients, 15.8% were immunocompromised, and 55.8% received antibiotic therapy upon admission. The mean IGS II score was 45.7, and the in-department mortality rate was 16.7%. Exposure to invasive devices is common: intubation (60.0%), central venous catheter (63.3%), and urinary catheter (85.3%). Among the 68,581 patients, 9.99% developed at least one monitored infection. The most frequently isolated microorganisms were P. aeruginosa (15.0%), S. aureus (11.2%), and S. epidermidis (8.4%). Since 2004, antibiotic resistance has been decreasing for S. aureus strains (15.8% MRSA in 2017). Resistance remains high for ESBL-producing E. coli (17.9% of ESBL-producing strains, with 1.5% resistant to imipenem), but appears to have stabilized in recent years. Incidence rates are 15.46 PNE per 1,000 intubation-days, 3.62 BAC per 1,000 hospitalization-days, 0.61 ILC, and 0.49 BLC per 1,000 CVC-days. These rates vary significantly from one department to another depending on patient characteristics. Over the past five years (2013 to 2017) across the entire network, certain risk factors have changed (increase in trauma patients, decrease in antibiotic therapy at admission, fewer elective and emergency surgeries), and the exposure ratio to intubation has decreased. We observe a decrease in incidence rates per 1,000 days of exposure: significant for BLC (-19.7%) and ILC (-21.8%). The multivariate analysis highlights a significant increase from 2014 to 2017 in PNE related to intubation (adjusted OR: 1.16; 95% CI: 1.09–1.22). However, it confirms a significant decrease in BLCs in 2017 (adjusted OR: 0.63; 95% CI: 0.49–0.81), which can be linked to improvements in professional practices related to invasive procedures in intensive care. With participation representing 47.2% of intensive care beds in France, the REA-Raisin data serve as a national benchmark for better understanding adverse events in intensive care and enable participating units to compare, evaluate, and guide their prevention efforts.