Distraction osteogenesis is an effective technique for reconstruction of the congenitally deficient mandible. However, the age at which it is best performed remains under discussion. Distraction performed at an early age, while possibly allowing the face to develop with a more normal functional matrix, may entail a higher rate of complications. Additionally, it is possible that subsequent asymmetric growth of the mandible may necessitate serial distraction. To address this issue, the clinical records and cephalometric radiographs of all patients less than 48 months of age undergoing mandibular distraction at New York University Medical Center between August of 1989 and August of 1997 were examined. There was a total of 14 patients ranging in age from 19 months to 43 months. Nine patients had a diagnosis of unilateral craniofacial microsomia, three had Treacher Collins syndrome, one had Nager syndrome, and one had bilateral developmental micrognathia. The average amount of distraction was 27 mm (range, 23 to 39 mm) in unilateral cases and 24 mm in bilateral cases (range, 15 to 31 mm). The period of clinical follow-up averaged 32.6 months (range, 12 to 92 months). All patients showed significant improvement in craniofacial appearance, and in four patients, long-term tracheostomy tubes were removed. There were two major complications. In one patient with craniofacial microsomia, there was a relapse in the early postretention phase related to the presence of a dentigerous cyst. This required removal of the cyst and repeat distraction. In the patient with Nager syndrome, a coronoid ankylosis developed requiring surgical release. There were no other major complications. The scars required revision in only two of the patients. Cephalometric analysis of the patients in the study revealed a differential in the rate of growth between the affected and the unaffected side in all cases of craniofacial microsomia. The affected side always grew at a slower rate than the contralateral side after the distraction process was complete. This led to a progressive asymmetry of the rami, clinically expressed by some degree of facial asymmetry and an occlusal cant. For this reason, secondary distraction was required in one patient and is planned in a second. Initial overcorrection of the patient would seem to minimize the likelihood that secondary distraction will be necessary. Distraction osteogenesis for reconstruction of the mandible in this subset of young patients was a safe and effective technique for improving the craniofacial skeletal form and appearance, with minimal associated morbidity. Longer follow-up is necessary to assess the full impact of growth in these cases.