A computer-implemented method for virtually representing an orthodontic treatment outcome using automated detection of facial and dental reference objects is disclosed herein. The computer-implemented method may involve modeling initial positions of an orthodontic patient's teeth in a three-dimensional (3D) virtual model and determining the effect of application of various force systems applied through an orthodontic treatment plan on the patient's teeth. Images of the patient's face may be used to identify facial reference objects. Dental reference objects of the patient's dentition may be identified from the 3D virtual model. A relationship between the facial reference objects and the dental reference objects may form the basis of modifications to the orthodontic treatment plan and/or modeling final orthodontic positions of the orthodontic treatment plan.