Under the auspices of the Network for Alert, Investigation, and Surveillance of Nosocomial Infections (Raisin) and in collaboration with the Study Group on Healthcare Workers’ Risk of Exposure to Infectious Agents (Geres), methods for monitoring blood exposure incidents (BEIs) have been the subject of a consensus and a national network since 2002. Each facility voluntarily, anonymously, and in a standardized manner documented all BEIs involving a staff member (including students and interns) reported to the occupational physician from January 1 to December 31, 2003. Data were collected on a Geres-adapted form documenting the circumstances of the blood exposure incident (nature, mechanism, equipment involved), its follow-up (immediate care, follow-up, and any prophylaxis), and the infectious status of the source patient. The incidence of blood exposure incidents was reported per number of hospital beds. In 2003, 6,316 blood exposure incidents were recorded in 228 facilities. The incidence of blood exposure incidents per 100 hospital beds was 7.5 (6.9 in 2002). Based on the 471,521 hospital beds recorded in France (SAE 2001 data), this allows for an estimate of 35,364 (32,423 in 2002) (95% CI: 35,034–35,741) the number of AES that were reported in 2003 to occupational physicians at French healthcare facilities. Even though the incidence data covers a slightly larger cohort of facilities than in 2002, the frequency of AES appears relatively similar. The 2003 results show a trend toward improved knowledge of the source patient’s status, with 16.9% of unknown HIV serology results compared to 20.3% in 2002, and 20.6% of unknown HCV serology results compared to 25.6% in 2002. While these rates remain too high, they may signal the beginning of the impact of awareness campaigns on this issue. Injection is the most common procedure involved, accounting for 19.2% of AEs, and in 84.4% (80% in 2002) of cases, it involves a subcutaneous injection. When examining the issue by type of equipment involved, there were 547 subcutaneous needles (10% of AES with at least one specified piece of equipment), to which must be added 474 insulin pens (8.7%) and 256 pre-filled heparin syringes (4.7%) (3.9% in 2002). With an AES rate of 41.5 per 100,000 procedures, care involving implanted catheter lines remains a challenge. While improvements in material quality have reduced the rebound effect when needles are withdrawn, ensuring the safety of these procedures must be a priority. The consolidation of 2003 adverse event surveillance data from various French networks once again highlights the significance of this phenomenon and the need for its prevention. It encourages the continuation of certain initiatives launched based on the 2002 report and opens up new avenues for consideration. (R.A.)