The aim of this study was to compare different treatment modalities for different severities of pediatric radial neck fractures in a large cohort of patients and determine prognostic factors. A retrospective, comparative study was conducted of all children treated for different severities of radial neck fractures at two level-1 pediatric trauma centers between 1990 and 2007. Pertinent data were collected, and the outcome was measured by the Tibone criteria. Several variables were compared to determine the prognostic value. Seventy-eight children were identified, 35 male and 43 female, at an average age of 7.8 years (range 3–15 years). Twenty-nine of the 78 (37%) children had associated injuries. Fracture patterns varied in their severity. Nineteen patients needed casting alone; 16 required closed reduction; of the children requiring surgery, 26 responded to percutaneous reduction and 17 were severe enough to need open reduction. Complications occurred in 28/78 (36%) children, 24 of which were related to stiffness. Seventeen of the 19 (89%) fractures that were amenable to casting alone had excellent or good outcomes, as did 11/16 (69%) in the closed reduction group. Among children treated operatively, 19/26 (73%) patients who underwent percutaneous reduction had an excellent or good outcome. Only 6/17 (35%) of the severe cases who underwent open reduction had an excellent or good outcome. The treatment method, essentially dictated by the severity of the cases, did correlate with the outcome (P=0.001). Compared with more severe fractures that required operative treatment, patients treated nonoperatively had a higher rate of excellent to good outcome (P=0.018). In particular, patients who underwent percutaneous reduction alone had improved outcomes versus open reduction (P=0.008). The outcome was not related to the presence of an associated injury (P=0.302). Initial fracture severity, as graded by the Judet classification, correlated with the outcome (P=0.004). Furthermore, age also expressed significance, as patients younger than 10 years of age tended to do better than those above 10 years of age (P=0.025). Children with less severe radial neck fractures amenable to nonoperative treatment fare well. Operative treatment of these fractures, particularly in severe cases when an open (not percutaneous) reduction is needed, is associated with a higher risk of poor outcome. Children over 10 years of age, especially those with more severe initial angulation, tend to fare worse. The presence of associated injuries does not appear to increase the risk of a poor outcome after treatment. There is a high rate of complications (36%), stiffness being the most common (24). This large bi-center retrospective study from two level-1 pediatric trauma centers emphasizes that in severe cases of radial neck fractures, open treatment does not necessarily decrease the risk of a poor outcome.