In 1998 the New South Wales (NSW) Health Department funded the development and implementation of the State's first standardized methodology for the surveillance of healthcare-associated infection for public hospitals. Fifteen pilot hospitals targeted inpatient groups considered to represent their core patient groups to act as sentinel measurements of patient safety. The aggregated rates of surgical site infection for coronary artery bypass graft (CABG) (chest & leg) surgery was 1.7% (95%CI: 1.1-2.5), CABG (chest only) 2.1% (95%CI: 1.0-3.7), vascular 7.1% (95%CI: 4.6-10.3), hip prosthesis 1.3% (95%CI: 0.5-2.7), knee prosthesis 6.1% (95%CI: 2.8-11.2) and colorectal 12.5% (95%CI: 9.5-16.1). The development of a bloodstream infection (BSI) associated with a central venous catheter (CVC) was not significantly (P=0.6) different when examined by duration of exposure with 3.7 BSI per 1000 line-days for CVC in situ six or more days compared with 4.0 BSI per 1000 line-days for CVC in situ for five or less days. A significantly (P<0.0001) greater proportion of patients whose CVC was in situ six or more days (6.8 per 100 patients, 95%CI: 4.2-10.2) developed a BSI compared with the proportion of patients whose CVC was in situ for five or fewer days (0.6 per 100 patients, 95%CI: 0.3-1.3). Significantly (P<0.0001) different rates of patients acquiring a new methicillin-resistant Staphylococcus aureus infection were found when hospital type was examined with rates ranging from 0.2 to 5.0 per 10000 occupied acute-care bed-days. The pilot highlighted that the collection of data for aggregation of some procedures and intravascular catheters may take many years before a reliable benchmark can be identified and many hospitals may not achieve reliable local rates annually. For surveillance to provide timely measures of patient safety we should consider surveillance methods for many small to medium sized hospitals that includes active surveillance only for infections with concurrent passive surveillance of the relevant denominators.