Aim: To increase the efficiency of diagnosing diseases of the temporomandibular joint by assessing a set of condylographic indicators and developing prognostic algorithms for the course and an optimized program of dental care for patients. Materials and Methods: A clinical retrospective cross-sectional comparative study was conducted with the participation of 470 patients with temporoman dibular joint pathology aged 18–76 (33 ± 12.9) years. The comprehensive examination included palpation of the masticatory muscles and temporomandibular joint, condylography, computed tomography, teleradiography, occlusiography, analysis of mandibular mobility, modeling of the jaws in the articulator, as well as logical and statistical data processing using standard methods (p<0.05). Results: The analysis of diagnostic and therapeutic approaches to patients with temporomandibular joint diseases allowed us to develop mathematical prognostic models and propose an examination program that takes into account the specifics of the development and course of the pathology. The algorithms for predicting the course of temporomandibular joint diseases based on complex clinical, condylographic and articulation methods were substantiated, which allowed to improve the dental care of patients. Conclusions: A modified program for diagnosing patients with the use of mathematical modeling (Wald’s method of sequential analysis) was proposed to assess the probability of progression of temporomandibular joint diseases. Further research will focus on expanding the primary data base, improving mathematical models, testing individualized prognostic algorithms and evaluating their clinical effectiveness