Since 2002, a national surveillance protocol for multidrug-resistant bacteria (BMR-Raisin) has enabled the five CClin (Centers for Coordination and Control of Nosocomial Infections) to generate, through their networks, standardized data on the two types of multidrug-resistant bacteria targeted by the national prevention program: methicillin-resistant Staphylococcus aureus (MRSA) and extended-spectrum beta-lactamase-producing Enterobacteriaceae (ESBL-producing Enterobacteriaceae). The selected indicators are the incidence rate per 1,000 hospital days (HD) (primary indicator) and, for short-stay patients, the attack rate per 100 hospitalizations (secondary indicator). Surveillance includes strains isolated from diagnostic specimens. Data generated by the CClin are aggregated at the national level. In 2003, 488 healthcare facilities (including 11% university hospitals, 39% public hospitals, 8% public psychiatric hospitals, and 18% private medical-care facilities) participated in the surveillance. These facilities accounted for nearly 160,000 beds, representing one-third of all beds in France and one-tenth of private beds. The incidence rate per 1,000 patient-days was 0.68 for MRSA and 0.14 for EBLSE overall. The incidence rate per 1,000 patient-days for MRSA was fairly consistent across regions (median 0.67, range 0.41–0.87), particularly in short-stay care (overall value 0.85, median 0.84, range 0.54–1.08) and in intensive care (overall rate 2.34, median 2.10, range 1.37–2.99). The overall rate was 0.40 in SSR-SLD. MRSA cases were distributed (a) by department type as follows: 39% internal medicine, 19% surgery, 11% intensive care, and 23% SSR-SLD; and (b) by specimen type as follows: 8% blood cultures, 8% serous and deep pus, 18% respiratory specimens, and 19% urine. The incidence rate per 1,000 patient-days for EBLSE varied widely by subregion (range 0.03–0.24) and was twice as high in short-stay care (0.18) as in SSR-SLD (0.09). Cases of EBLSE were distributed (a) by department as follows: 36% in internal medicine, 16% in surgery, 15% in intensive care, and 24% in SSRSLD; and (b) by sample type as follows: 5% blood cultures, 5% serous and deep pus, 14% respiratory specimens, and 61% urine. By extrapolation, based on these data and the number of hospital days in France, the number of MRSA cases isolated from diagnostic specimens can be estimated at approximately 68,000 per year in France in 2003, including 5,000 bacteremias. The overall incidence rate of MRSA and VRE did not decrease between 2002, the first year of BMR-Raisin surveillance, and 2003 (0.63 and 0.68 MRSA per 1,000 hospital days and 0.13 and 0.14 VRE per 1,000 hospital days, respectively). However, a slight decrease in incidence was observed in the intensive care unit. (R.A.)