Rare, but serious, the possible accidents of medical or surgical procedures without fault compromise the doctor patient relationship. A pro-compensation philosophy leads to a drift in the classical concept of prejudicial act. This type of confusion adds a random/aleatory legal risk to the random/aleatory medical risk. Determining the limit between random/aleatory risk and malpractice is one of the difficulties in medical evaluation of responsibility. In order to evaluate the random/aleatory risk without malpractice, in relationship to the medical team, we propose a notation grid for the facts of the case, based on simple, precise, identifiable criteria present in all medical records and accepted by both the practising physicians and medico legal experts. This medico-legal research was done in a double blind study of 45 cases which intially brought up a random/aleatory risk related problem for three medical procedures: coronary arteriography, colonoscopy and laparascopy. In 66.6% of the cases the random/aleatory risk was ruled unrelated to any malpractice, but in 33.4% of the cases observed, the initial random/aleatory risk was transformed into actionable malpractice, essentially due to faulty surveillance or tardy treatment of the complication. This notation system, carried out in identical and well defined terms, should allow a reasonable comparison of cases for medico-legal research or computerised databases. It is reserved exclusively for use by medical experts or consultants so they can give a reasoned technical opinion in order to assist the judge in his legal decision. The objective of the proposed grid is to homogenise the concepts used by experts in all national courts. After a multicenter validation, this grid could have an European future as to the boundary between random/aleatory risks and medical malpractice and responsibility.