AbstractA body of evidence has been accumulated to demonstrate that current practice is not sufficiently safe for several stages of central laboratory testing. In particular, while analytical and perianalytical steps that take place within the laboratory are subjected to quality control procedures, this is not the case for several pre- and post-analytical steps. The ubiquitous application of auto-identification technology seems to represent a valuable tool for reducing error rates. A series of projects in Oxford has attempted to improve processes which support several areas of laboratory medicine, including point-of-care testing, blood transfusion, delivery and interpretation of reports, and support of decision-making by clinicians. The key tools are auto-identification, Internet communication technology, process re-engineering, and knowledge management.