Orthodontic treatment acts through the application of forces and/or by stimulating and redirecting the functional forces within the craniofacial complex. Considering the interrelationship between craniomandibular and craniocervical systems, this intervention may alter craniocervical posture. Thus, our aim is to (a) compare craniocervical posture, hyoid bone position, and craniofacial morphology before, after, and also in the contention phase at least one year after the orthodontic treatment, in patients with temporomandibular disorders and (b) to verify whether the presence of condylar displacement, the skeletal class, or the facial biotype interferes with the abovementioned outcomes. To do so an observational, analytical, longitudinal, and retrospective design study was carried out. A non-probabilistic convenience sampling method was applied. The sample consisted of clinical records of patients diagnosed with temporomandibular disorders in order to compare pre-orthodontic treatment with post-orthodontic treatment (n = 42) and contention phase data (n = 26). A cephalometric analysis of several variables was performed. The p-value was set as 0.05. When the pre- and post-orthodontic treatment data were analyzed, there were statistically significant changes in variables concerning craniocervical posture (CV angle, C0-C1, and AA-PNS) and also concerning hyoid bone position (C3-Rgn). When pre- and post-orthodontic treatment and contention phase data were analyzed the variables concerning craniocervical posture (C0-C1, CVT/Ver, NSL/OPT, NSL/CVT, NSL/Ver; OPT/CVT, OPT/Ver) and facial biotype had statistically significant changes. This allowed us to conclude that in the sample studied, there were significant differences regarding hyoid bone position (pre- versus post-orthodontic treatment) and craniocervical posture (pre- versus post-orthodontic versus contention), with the craniocervical posture being prone to return to basal values. The presence of condylar displacement was found to significantly increase the H-H1 distance in the three moments of evaluation. Facial biotype was found to significantly increase the NSL/Ver angle on hypodivergent compared to hyperdivergent in the contention phase.