To determine the factors that influenced doctors' prioritization and decisions on safe waiting time for coronary artery bypass surgery, 50 'paper patients', based on a random sample of cases who actually had surgery, were assessed by 33 clinicians. We used linear regression models to reflect the impact of clinical and non-clinical 'cues' on safe waiting time and priority decisions. The benefits of surgery tended to be over-estimated. For example, the average perceived gain in life expectancy for patients with left main-stem disease was 6.74 years. However, models incorporating only the perceptions of benefit as independent variables (i.e. the anticipated symptom reduction, MI risk reduction and life expectancy extension), had only modest explanatory power (mean R2 was 0.55 for safe waiting time, and 0.56 for priority decisions). Models which incorporated perceptions of benefit and the cases' clinical and non-clinical characteristics had generally much higher explanatory power (mean R2, 0.83 and 0.86, respectively). Lifestyle and demographic variables had much less impact on the doctors' judgements than the major clinical cues of angina severity and left main-stem stenosis. Demographic and lifestyle cues had different impacts on safe waiting time and priority for about 25% of doctors.