The most current clinical definition of anaphylaxis is that proposed by Sampson et al, and different scoring systems have been proposed to assess its severity. 1,2 Anaphylaxis symptoms range from mild, self-limiting local reactions to life-threatening anaphylaxis.All anaphylaxis guidelines emphasize early recognition and the need for prompt intramuscular adrenaline injection as the first-line treatment. 3 However, adrenaline for anaphylaxis is underused by patients, families and even by health professionals. In the European anaphylaxis registry, 27% of patients treated by a health professional received adrenaline. 4 Given this low rate, spontaneous resolution of anaphylaxis symptoms may occur without adrenaline injection. There are few data on the course of anaphylaxis manifestations in children treated with adrenaline or not. This study aimed to describe the time course of clinical manifestations in children experiencing anaphylaxis. All children (≤15 years) admitted for anaphylaxis to one of 18 paediatric emergency care units in the Nord-Pas-de-Calais region of France were prospectively included (2015-2017). A standardized questionnaire was filled out using data collected from the family and the medical records. Patients were diagnosed with anaphylaxis if they fulfilled the criteria reported by Sampson et al. 1 All medical records were analysed and reviewed by three independent physicians including one anaphylaxis expert to confirm the diagnosis. The 5-grade Astier score was used to analyse anaphylaxis severity. 2,5 (Figure S1). We assessed the effect of baseline characteristics on the risk of progression (improvement vs worsening) of anaphylaxis severity by using Chi-square and Student's t tests. Association of Astier score with adrenaline use was investigated by Mann-Whitney U test. Statistical testing was conducted at the two-tailed α-level of 0.05. This study was approved by the French Data Protection Authority and the Advisory Committee for Information Processing in Health Research.A total of 152 cases of anaphylaxis were identified, but three families refused the inclusion. The final sample consisted of 149 cases; each child had been admitted only once. Mean age at inclusion was 7.4 (4.7) years, and 87 (58%) were boys. Thirty-five (23%) children had a known food allergy to the trigger involved in the study case, and 17 (20%) a history of past anaphylaxis. An emergency kit had been prescribed for 35 children, including an adrenaline autoinjector (AAI) in 31.The main anaphylaxis triggers were foods (n = 110; 74%, including peanut, n = 32; 21%). The other triggers were medication (n = 13; 9%) and hymenoptera venom (n = 1). Twenty-five (17%) reactions were considered as idiopathic. Cofactors were involved in 35 (23%) cases including viral infection (n = 23; 15%) and physical exercise (n = 11; 7%).The emergency medical services (EMS) were called by the family in 63 (43%) cases, and 95 (63%) patients went directly to the emergency care unit. In all, 55 (37%) children received prehospital care by Guillaume Poues...