Aims: To determine the effect of a simple, user-friendly algorithm developed as a guidance for reducing the need for allogeneic blood transfusions in total hip and knee arthroplasty. Methods: Allogeneic blood use in elective primary hip and knee arthroplasty was prospectively determined in an observational study, with 229 consecutive patients assessed before (November 1998 to September 1999) and 262 consecutive patients assessed after the implementation of a simple one-page algorithm in a Swiss orthopaedic tertiary care centre (October 1999 to October 2000). The algorithm was widely disseminated and placed in all charts of patients undergoing total hip or knee arthroplasty and handed out to any of the orthopaedic centreñs physicians and nurses from October 1st 1999 onwards. The medical staff was required to follow the algorithm for all medical decision making related to allogeneic blood transfusion. Results: Patients before and after the implementation of the algorithm were well matched. The mean number of units of allogeneic blood transfused per patient was reduced from 0.88 units to 0.43 units (p< 0.0001 by Mann-Whitney test). 80 out of 229 patients received allogeneic blood transfusions before, and 49 out of 262 after implementation of the algorithm (odds ratio 0.43, 95% CI 0.28 to 0.65, p< 0.001). There was also a highly signiþcant decrease in the mean number of units of autologous blood donations, from 0.56 to 0.13 units (p< 0.0001). There was only one ischaemic event in each group. Conclusions: Allogenic blood transfusions in primary hip and knee arthroplasty may be reduced by implementing a simple user-friendly algorithm. Our results need to be conþrmed in a large-scale cluster-randomised trial.