The aim of this pilot study was to investigate anaesthetists' assessment of the ability of patients to increase cardiac output over a range of clinical scenarios and of their perceived 'likelihood of transfusion' in these scenarios. Specialist anaesthetists were given a questionnaire with clinical cues in the form of diagnoses about theoretical patients. They were asked to use 100 mm visual analogue scales (VAS) for their assessments of each patient's cardiac reserve and their 'likelihood of transfusion' of these patients; the endpoints of the VAS being 'Very low' (0 mm) to 'High' (100 mm), and 'Do not transfuse' (0 mm) to 'Transfuse' (100 mm) respectively. The assessment of patients' cardiac output reserve by anaesthetists (n = 54) showed great variation; for example, a patient with severe aortic stenosis was perceived overall to have a limited ability to increase cardiac output (mean VAS 16 mm) but there was considerable variation between anaesthetists (25-75 percentiles 10 mm to 21 mm). Assessment of 'likelihood of transfusion' (n = 42) also had great variation; as an example a patient with 'angina' with a haemoglobin of 95 g l(-1) was perceived overall to have an average likelihood of transfusion of 50 mm, but the 25-75 percentiles ranged from 33 mm to 71 mm. This study suggests that inter-anaesthetist variability in the assessment of a patient's 'cardiac output reserve' and his 'likelihood of transfusion' is large.