OBJECTIVE: This article describes a method for investigating attitudes towards prioritisation in medicine. SETTING: University of Kuopio, Finland. DESIGN: The method consisted of a set of 24 paired scenarios, which were imaginary patient cases, each containing three different ethical indicators randomly selected from a list of indicators (for example, child, rich patient, severe disease etc.). The scenarios were grouped into 12 random pairs and the procedure was repeated four times, resulting in 12 scenario pairs arranged randomly in five different sets. SURVEY: This method was tested with four groups of subjects (n = 8, n = 47, n = 104 and n = 36). RESULTS: Children and patients with a severe disease were prioritised in all groups. The aged, patients with a mild disease and patients with a self-acquired disease were negatively prioritised in all groups. Poor or rich patients were prioritised in some groups but negatively prioritised in others. CONCLUSIONS: The validity and reliability of this method are good and it is suitable for investigating attitudes towards medical prioritisation.