The aim of this survey was to evaluate anaesthesiologists' ability to quantify a given risk of perioperative morbidity. Descriptive study. Nine hypothetical patients with well-defined single-organ failure related to three validated perioperative risk indexes (cardiac risk, respiratory-failure risk and cirrhosis decompensation) were described in vignettes and submitted to 30 anesthesiologists issued from two institutions and distributed in two groups (10 residents and 20 staff anesthesiologists). Physicians were asked to estimate the patient's risk using qualitative (scored using a categorical scale) and quantitative (scored using a number, i.e., the estimated frequency that a given complication would occur in the given case) evaluations of the considered risk for each case proposed. Both qualitative and quantitative evaluation of the most severe cases showed a tendency to underestimate respiratory and hepatic risks and a tendency to overestimate cardiac risk. No major difference in the accuracy of evaluations was found between trainees and staff anaesthesiologists. Both provided imprecise risk evaluation and had difficulty to define the expected level of risk, even for clear-cut clinical situations described in the vignettes. Both qualitative and quantitative evaluations were overall very imprecise. This survey showed that a rigorous evaluation of a patient's perioperative risk cannot be considered as intuitive and needs objective material supports to be performed.