Introduction: The aims of this research were to test the repeatability and reproducibility of measurements obtained by different molar landmarks and to determine whether the treatment changes measured by them are comparable. Materials: Pretreatment (T0) and postdistalization (T1) lateral cephalograms of 40 patients with a Class II molar relationship were used. Using the mesial cusp tip, distal crown contour, and centroid, 6 linear and 3 angular measurements were evaluated. The intraexaminer and interexaminer reliabilities were assessed with the Bland-Altman method. The T0 to T1 differences were compared by analysis of variance, followed by multiple comparisons with the Bonferroni adjustment. Results: The greatest intraexaminer agreement intervals were correlated with centroid for tipping and vertical movement, and with distal crown contour for anteroposterior movement at T0 and T1. The narrowest interexaminer agreement intervals were observed with mesial cusp tip for all measurements and comparisons at both time points, except for the anteroposterior measurements at T1. Statistical significance was observed between the measurements made with mesial cusp tip and distal crown contour for tipping and with mesial cusp tip and centroid for distalization. For vertical movement, the treatment changes were statistically significantly greater with distal crown contour. Conclusions: Mesial cusp tip was the most repeatable and reproducible landmark. Clinically unimportant differences were found in some comparisons of treatment changes obtained by different landmarks.