Combined allergological and anaesthetic consultations have been started in the last few years in eight French Teaching Hospitals so as to explore peranaesthetic anaphylactoid shocks. A survey was carried out in these centers in order to collect patients investigated with the same protocol, for the assessment of the incidence of anaphylaxis in France, as well as the involved drugs. Investigations were always carried out at least 6 to 8 weeks after the accident. The tests used to diagnose IgE-dependent anaphylaxis were skin tests (prick and intradermal tests, carried out in all eight centers), the radioimmunological assay of specific anti-quaternary ammonium IgE, together with an inhibition test with thiopentone and propofol (six centers), leukocyte histamine release (five centers) and human basophil degranulation tests (three centers) for those drugs for which no specific antibody assay exists. The collected data involved 1,240 patients, investigated within the last four years. Anaphylaxis was diagnosed in 821 patients (66.2%). Muscle relaxants were responsible in 668 cases (80% of cases of anaphylaxis). Suxamethonium was the main cause (54.3% of shocks due to muscle relaxants), followed by vecuronium (15.3%). General anaesthetics (hypnotics and benzodiazepines) were responsible for 9.2% of all cases of anaphylaxis opioids for 2.6%. There were only three cases of shock due to local anaesthetic agents. Latex and ethylene oxide are becoming increasingly involved. It would therefore seem mandatory to carry out after any anaphylactoid accident an assessment with sensitive and specific tests for anaphylaxis. Diagnosing anaphylaxis means that the involved drug should be used never again in that patient. Because muscle relaxants are by far the most involved drugs, anaesthetists should use them only when really required.