Background: CBCT is increasingly used to assess alveolar bone in orthodontic patients; however, reliability and agreement for detecting changes in follow-up assessments remain uncertain. This study evaluated the reliability and agreement of linear marginal alveolar bone level measurements, and agreement in detecting fenestrations in CBCT volumes acquired from different units before and after active orthodontic treatment in adolescents, using a multiplanar reconstruction-based measurement protocol for follow-up studies. Methods: CBCT volumes (n = 96) from 48 adolescents (mean age 14.4 years) enrolled in a randomised controlled trial on fixed-appliance treatment for crowding were acquired at baseline and posttreatment using five distinct CBCT units after objective image-quality testing. Buccal and lingual bone levels at the central incisors, canines, and second premolars were measured from the cementoenamel junction to the alveolar crest according to a predefined protocol. Three independent raters performed linear measurements, two assessed fenestrations greater than 2.2 mm, and one repeated all measurements. Reliability was evaluated using intraclass correlation coefficients. Agreement for the bone level measurements was assessed using Dahlberg’s formula and visualised with Bland–Altman plots. Minimal detectable change was calculated per site, and percentage agreement was used for fenestration detection. Results: Intrarater reliability ranged from moderate to excellent at both time points. Dahlberg’s random errors ranged from 0.19 to 0.71 mm at baseline, and 0.19 to 1.07 mm posttreatment. Interrater reliability ranged from poor to good, with random errors ranging from 0.23 to 2.54 mm. Bland–Altman plots revealed minor systematic bias for the intrarater measurements. Minimal detectable changes demonstrated the ability to detect clinically relevant alveolar bone changes. Agreement for fenestration detection was high (intrarater ≥ 81%, interrater ≥ 74%). Conclusion: The protocol demonstrated intrarater reliability and agreement suitable for assessing clinically relevant changes in marginal bone levels and for detecting fenestrations in follow-up studies of adolescents treated orthodontically for crowding without extraction. However, the results were influenced by rater variability and imaging conditions and were most reliable when the assessments were performed by the same rater. Assessment of reliability and agreement under study-specific conditions is recommended. The findings highlight both the potential and methodological challenges of CBCT-based alveolar bone assessments, offering implications for clinical practice and future research in this field. Trial registration: CROWDIT: ClinicalTrials.gov (NCT05664282).