Objective: To analyze the epidemiological characteristics of hospitalized children with severe burn from several regions in China during 3 years, so as to provide evidence for prevention of burns in children. Methods: Relying on the entry system of epidemiology data and biological sample of severe burn from multicenter in clinic, medical records of children with severe burn, aged 18 and under, hospitalized in 6 burn wards from February 2012 to February 2015 were collected. The children were divided into 5 age brackets: less than or equal to 1 year old, more than 1 year old and less than or equal to 3 years old, more than 3 years old and less than or equal to 6 years old, more than 6 years old and less than or equal to 12 years old, more than 12 years old and less than or equal to 18 years old. Meanwhile the children were divided into rural and urban children according to their residences. Data of gender and residence of children in the 5 age brackets, cause of injury, location of injury, total burn area, wound site, inhalation injury and combined injury, and source of hospitalization expenses of children in the 5 age brackets and two types of residence, and outcome and length of hospital stay of the children were analyzed. The cause of injury of children in different location of injury was analyzed. In addition, they were divided into 2 age brackets: less than or equal to 6 years old and more than 6 years old and less than or equal to 18 years old, and then incidences of hand and foot burn injury were compared. Data were processed with chi-square test, and the correlation between age and total burn area was processed with Spearman correlation analysis. Results: Four hundred and forty out of 1 106 inpatients with severe burn were children, accounting for 39.8% who were included in the system. (1) The majority of children were male (270, 61.4%). The number of children more than 1 year old and less than or equal to 3 years old ranked the first (222, 50.5%) in the 5 age brackets. The ratio of children from rural areas to that from urban areas was 2.9∶1.0. There were no statistically significant differences in both gender and residence of children among the 5 age brackets (with χ2 values respectively 7.649 and 9.399, P values above 0.05). (2) Scald was the most common cause of burn. There was statistically significant difference in injury cause of children among the 5 age brackets (χ2=136.307, P<0.001). There was no statistically significant difference in injury cause of children among the two types of residence (χ2=5.164, P>0.05). (3) Private house was the most common location of injury. There was statistically significant difference in location of injury of children among the 5 age brackets (χ2=124.930, P<0.001). There was no statistically significant difference in location of injury of children among the two types of residence (χ2=3.364, P>0.05). There was statistically significant difference in injury cause of children in different location of injury (χ2=118.284, P<0.001). (4) Most of children were with total burn areas from 10% to 39% total body surface area. There was statistically significant difference in total burn area of children among the 5 age brackets (χ2=103.568, P<0.001). There was positive correlation between age and total burn area (r=0.177, P<0.001). There was no statistically significant difference in total burn area of children among the two types of residence (χ2=16.213, P>0.05). (5) Trunk, lower extremity, and upper extremity were the most common wound sites, respectively. There was statistically significant difference in wound site of children among the 5 age brackets (χ2=45.674, P=0.019). There was statistically significant difference in incidence of hand and foot burn between children less than or equal to 6 years old and children more than 6 years old and less than or equal to 18 years old (with χ2 values respectively 29.188 and 14.612, P values below 0.01). There was no statistically significant difference in wound site of children among the two types of residence (χ2=8.515, P>0.05). (6) Twenty-seven children suffered inhalation injury. The main age bracket was more than 12 years old and less than or equal to 18 years old (8 children). The main residence was rural area (18 children). The main cause of inhalation injury was flame burn (23 children). Nine children suffered combined injury, among which the children more than 12 years old and less than or equal to 18 years old accounted for the highest ratio (5 children), and the urban children accounted for higher ratio (5 children). (7) Among the 437 children, most of their hospitalization expenses were at their own expense. There was statistically significant difference in the source of hospitalization expenses of children among the 5 age brackets (χ2=17.917, P=0.001). There was no statistically significant difference in the source of hospitalization expenses of children among the two types of residence (χ2=0.749, P>0.05). (8) Among the 437 children, 34 children abandoned treatment and were discharged from hospital, attributed to lack of funding. Seventy-eight children were discharged with a better health condition and 347 were cured. The condition of 6 children worsened and 6 children died. Mean length of hospital stay was 28.6 days for all the children, and 8.8 days for the deteriorated and dead children. Conclusions: Children were the major group of patients with severe burn in China. Male children less than or equal to 6 years old were common with scald as the major cause of injury, private house as the major location of injury, and trunk, lower and upper extremity as the most common wound sites, their own expenses as the major source of hospitalization expenses. There were statistically significant differences in cause of injury, location of injury, total burn area, wound site, and hospitalization expenses source of children among the 5 age brackets.