Interventions to change physician transfusion behavior are often evaluated by examining the amount of red blood cell (RBC) units transfused or the proportion of patients transfused before and after the intervention. The pre-transfusion haemoglobin concentration is a sensitive measure of transfusion practice, but has not been used to evaluate behavioral interventions. We examined the effect of a Danish National Board of Health December 2007 transfusion guideline on the behavior of clinicians treating acute myeloid leukaemia (AML). We compared the effect of the guideline on pre-transfusion haemoglobin concentrations with other measures of transfusion behavior, including use of RBC units and proportion of patients transfused. No change in transfusion behavior could be demonstrated by examining amount of RBC units transfused and proportion of patients transfused. Conversely, the pre-transfusion haemoglobin concentration fell significantly. Pre-transfusion haemoglobin determination is a sensitive measure of the effect of an intervention to change physician transfusion behaviour.