This study compared facial growth in three groups of patients with bilateral complete cleft lip/palate: those who had (1) no premaxillary osteotomy, (2) premaxillary osteotomy before age 8 years, and (3) premaxillary osteotomy after age 8 years. Of 24 children with bilateral complete cleft lip/palate, 7 had early premaxillary osteotomy (mean age, 6.1; range, 3.7 to 7.6 years), 10 had late osteotomy (mean age, 11.2; range, 8.3 to 20.7 years), and 7 did not require premaxillary repositioning and served as controls (mean age, 12.4; range, 6.4 to 17.8 years). Presurgical and postsurgical lateral cephalograms were digitized using the Dentofacial Planner software; most current lateral cephalograms comprised the control group. Forty-one bony and 25 soft-tissue landmarks were digitized, and 8 angles were measured: SNA, (sella-nasion-A point), SNPg (sella-nasion-pogonion), ANB (A point-nasion-B point), NAPg (nasion-A point-pogonion), ST convexity (glabella-subnasale-soft-tissue pogonion), Sn-G vertical (line perpendicular to the horizontal plane dropped from glabella and distance measured from subnasale to this vertical), Cm-Sn-Ls (columella-subnasale-abial superioris), and Sn-Gn-C (subnasale-soft-tissue gnathion-chin point). Statistical difference in mean preoperative and postoperative values were measured with analysis of variance. Tests of significance were adjusted for multiple comparisons using the Bonferroni correction. Mean age at follow-up for early, late, and control groups was 11.8, 14.0, and 12.4 years, respectively. Mean follow-up for early and late groups was 5.7 and 2.8 years. There was a significant preoperative difference among the three groups for mean SNA (p < 0.01), ANB (p < 0.01), and NAPg (p < 0.01). Bonferroni analyses revealed that the early group had significantly greater SNA, ANB, and NAPg angles than the late (p < 0.01) and control groups (p < 0.05). There was a significant postoperative difference among groups for ANB (p < 0.05); Bonferroni analyses also showed that the control group had a significantly greater ANB than the late group (p < 0.05). The t test for equity of means established postoperative change for SNA (p < 0.01), ANB (p< 0.01), NAPg (p < 0.01), and ST convexity (p < 0.01) for the early group was significantly greater than for the late group. Children who required early premaxillary positioning had more significant preoperative deformity; however, this group's postoperative profile was not, on average, significantly different from either the late or control groups. Our findings that the early group had more significant change with premaxillary osteotomy than the late group suggest that premaxillary positioning can be done before completion of facial growth without compromise.