Background: the incidence of perioperative cardiac arrest and mortality in children is higher than in adults. This survey evaluated the incidence, causes, and outcome of perioperative cardiac arrests in a pediatric surgical population in a tertiary teaching hospital between 1996 and 2004.Methods: the incidence of cardiac arrest during anesthesia was identified from an anesthesia database. During the study period, 15 253 anesthetics were performed in children. Data collected included patient demographics, surgical procedures (elective, urgent, or emergency), ASA physical status classification, anesthesia provider information, type of surgery, surgical areas, and outcome. All cardiac arrests were reviewed and grouped by the cause of arrest and death into one of four groups: totally anesthesia-related, partially anesthesia-related, totally surgery-related, or totally child disease or condition-related.Results: There were 35 cardiac arrests (22.9 : 10 000) and 15 deaths (9.8 : 10 000). Major risk factors for cardiac arrest were neonates and children under 1 year of age (P < 0.05) with ASA III or poorer physical status (P < 0.05), in emergency surgery (P < 0.05), and general anesthesia (P < 0.05). Child disease/condition was the major cause of cardiac arrest or death (P < 0.05). There were seven cardiac arrests because of anesthesia (4.58 : 10 000) - four totally (2.62 : 10 000) and three partially related to anesthesia (1.96 : 10 000). There were no anesthesia attributable deaths reported. The main causes of anesthesia attributable cardiac arrest were respiratory events (71.5%) and medication-related events (28.5%).Conclusions: Perioperative cardiac arrests were relatively higher in neonates and infants than in older children with severe underlying disease and during emergency surgery. The fact that all anesthesia attributable cardiac arrests were related to airway management and medication administration is important in prevention strategies.