BACKGROUND Our aim was to assess effectiveness of the cervical vertebral maturation (CVM) method to predict circumpubertal craniofacial growth in the postpeak period. METHODS The CVM stage was determined in 176 subjects (51 adolescent boys and 125 adolescent girls) on cephalograms taken at the end of treatment (T2; mean ages, 15.75 years [boys] and 15.23 years [girls]) in subjects from the postretention database at the University of Washington in Seattle. Craniofacial growth was evaluated from the following measurements on cephalograms at T2 and end of follow-up (T3) (mean ages, 29.01 years [men] and 28.08 years [women]): condylion to gnathion, condylion to gonion, gonion to gnathion, sella to gnathion, nasion to menton, anterior nasal spine to menton, and sella to gonion. The change of each variable from T2 to T3 was assessed with paired t tests. Parametric (t tests or analysis of variance [ANOVA]) or nonparametric (Mann-Whitney or Kruskal-Wallis) tests were used to detect intergroup differences. RESULTS One hundred eight subjects (35 boys, 73 girls) demonstrated CVM stage 3, 56 (16 boys, 40 girls) were in CVM stage 4, and 12 (all girls) were in CVM stage 5 at T2. Intrasex comparisons showed that boys in CVM stages 3 and 4 could be differentiated regarding changes of all variables. In the girls, only those in CVM stages 3 and 4 could be differentiated based on the amount of changes of 2 measurements: condylion to gonion and sella to gonion. Intersex comparisons showed that boys in CVM stage 3 had significantly more changes than girls (P <0.01). Boys in CVM stage 4 showed significant differences compared with girls in CVM stage 4 for only 2 variables (sella to gonion and condylion to gonion; P <0.001 and P = 0.012, respectively). CONCLUSIONS The CVM method was modestly effective in determining the amount of postpeak circumpubertal craniofacial growth.