This study used a computer-based simulation of a patient with gastrointestinal symptoms for an investigation of decision-making amongst 122 general practitioners participating in the 6th Nordic Congress of General Practice. The simulation model, developed in the computer software COMCATS, makes use of an IBM-compatible personal computer attached to an automated slide-projector. All participants were given a short introduction to the patient and his symptoms, but all further information was optimal and selected by the doctor. Large variations were found between the participants with respect to clinical patterns: for example, expenses for further examinations ranged from 0 to 2,400 Danish crowns. For research into clinical decision-making the great advantages of the computer-model are that it allows for a standardisation of the clinical situation, gives the doctor an opportunity to select between optional information and permits an automatic gathering of huge amounts of information about the decision-making process.