In this cephalometric investigation, we compared the long-term effects of an initial phase of rapid maxillary expansion and facemask (RME/FM) therapy followed by comprehensive edgewise therapy with the effects of growth in untreated, matched controls. The treated sample consisted of 34 patients who underwent RME/FM treatment before the pubertal growth spurt (average age, 8 years 3 months at the beginning of treatment). At the final observation period (average age, 14 years 10 months), all patients were in decelerative growth phases as determined by the cervical vertebral maturation (CVM) method. After the first 10 months of active treatment, significant favorable changes in both the maxillary and the mandibular skeletal components were noted. The forward movement of the maxilla was 1.8 mm greater than in the controls, mandibular projection was reduced by almost 3 mm, and the relative sagittal intermaxillary discrepancy improved by 4.3 mm, as measured by the Wits appraisal. During the posttreatment period, the treated and untreated Class III subjects generally grew similarly, although the skeletal relationship of the maxilla to the mandible remained unchanged in the RME/FM group, whereas the controls had an increased skeletal discrepancy of 3.0 mm. Over the long term, there was a slightly greater increase in midfacial length (1.6 mm) in the treatment group than in the controls. Similarly, the distance from Point A to nasion perpendicular decreased by 1.2 mm in the treated group. The overall increase in mandibular length was 2.4 mm less in the RME/FM group than in the controls, and mandibular projection relative to nasion perpendicular was 3.0 mm less in the treated group. The change in the Wits appraisal was substantial between groups (6.1 mm), with an improvement in the intermaxillary relationship in the treated group (3.4 mm); the Wits appraisal worsened (-2.7 mm) in the untreated controls. No clinically significant differences were observed between the groups in the vertical dimension. Overjet increased significantly in the treated group relative to the controls (4.4 mm), whereas the molar relationship decreased significantly (-3.9 mm). It appears that the favorable skeletal change observed over the long term is due almost entirely to the orthopedic correction achieved during the RME/FM protocol. During the posttreatment period that includes the pubertal growth spurt, craniofacial growth in RME/FM patients is similar to that of untreated Class III controls. Aggressive over-correction of the Class III skeletal malocclusion, even toward a Class II occlusal relationship, appears to be advisable, with the establishment of positive overbite and overjet relationships essential to the long-term stability of the treatment outcome.