OBJECTIVE: This study aims to reveal the reconstruction process in pediatric patients with extracapsular condylar fractures after conservative treatment. We clarify that the "upright" position (or "recontouring" or favorable prognosis) of condyles is not a result of the anatomical reduction of the deviated condylar processes but originates from the remodeling of the skeleton. We also explore the related mechanism. METHODS: The sample consisted of 27 pediatric patients aged less than 12 years who presented with extracapsular condylar fractures and were treated conservatively within an 8-year period (June 2011-April 2019). Data on the age, gender, date of injury, mechanism of trauma, location and pattern of mandibular condylar fracture and associated injuries and treatment methods of the patients were obtained. The process of bone remodeling in condyles was also recorded and analyzed. RESULTS: The 27 children in this study sustained 33 extracapsular condylar fractures over the 8-year period of record retrieval. Amongst these fractures, 8 (24.2%) and 25 (75.8%) were condylar neck and condylar base fractures, respectively. Deviation and green-stick fractures were the predominant types and accounted for over 3 quarters of the condylar neck and base fractures (28, 84.8%), followed by dislocation fracture (3, 9.1%), displacement fracture (1, 3.0%), and non-displaced fracture (1, 3.0%). The period of follow-up ranged from 2 days to 257 days (average, 58.78 days). Only 1 patient with bilateral extracapsular condylar fractures showed vertically reconstructed condyles, which indicates an upright position of the condylar processes. One patient showed less angulation after treatment than before treatment, 1 patient revealed greater angulation after treatment than before treatment and all other patients (20 patients) showed the same angulation pre- and post-treatment. Both patients with only extracapsular condylar fractures showed no obvious deviations in dentition and facial asymmetry after their injury and treatment. The shortest and longest times observed for bone remodeling were 33 and 256 days, respectively. Children whose condylar head remained completely or at least partly inside the glenoid fossa showed satisfactory remodeling results during follow-up. Computed tomography scan during follow-up generally showed bone regeneration in the lateral condyle articular surface and the medial portion of the ascending ramus and bone resorption in the displaced direction (ie, the medial condyle head became sharp). Condylar heads displaced completely outside of the glenoid fossa showed serious shortening of the ascending ramus, and no obvious bone remodeling was observed. Only 1 patient with bilateral extracapsular condylar fractures showed a normal contour (ie, a vertically reconstructed condyle reflecting the upright position of the condylar processes) after 8 months. CONCLUSION: Stress stimulation originating from the glenoid fossa and ascending ramus of the mandible is a prerequisite for good condylar reconstruction. Conservative treatment could be carried out if the condylar head remains completely or at least partly inside the glenoid fossa. When the condylar head is dislocated completely outside the glenoid fossa, the glenoid-condylar relationship ceases to exist, joint function is lost and the height of the ascending ramus is significantly reduced. In this case, open reduction may be suitable.