Objective: To quantify the relative risk of multiple occlusal variables for muscular disorders of the stomatognathic system. Method and Materials: Thirteen occlusal features were clinically assessed by the same three operators: partial unilateral posterior reverse articulation, anterior open occlusal relationship, vertical anterior overlap (normal value 4 mm), complete unilateral posterior reverse articulation, anterior overjet (normal value 5 mm), incisor midline discrepancy (normal value 2 mm), Angle Class I canine and molar relationship; plus dynamic occlusion features such as length and symmetry of retruded contact position-intercanine position (RCP/ICP) slides (normal value 2 mm), occlusal guidance patterns, and mediotrusive and laterotrusive interferences. The sample consisted of 156 women with only muscular disorders according to the Group I Research Diagnostic Criteria for temporomandibular disorders and 82 healthy women (control group). A stepwise multiple logistic regression model was used to identify the significant correlation between occlusal features and disease. Results: The odds ratio for myofascial pain was 2.6 for absence of canine guidance, 2.0 for laterotrusive interference, 2.3 for mediotrusive interference, and 1.9 for reverse articulation. Other occlusal variables did not reveal statistical significance. The percentage of the total log likelihood for myofascial pain explained by the significant occlusal factors was acceptable with a Nagelkerke R2 = 0.2. The final model including the significative occlusal features revealed an optimal discriminant capacity to predict patients with myofascial pain with a sensitivity of 62.2%, or healthy subjects with a specificity of 93.6%, and an accuracy of 82.8%. Conclusions: Few occlusal features show a significative predictive value for myofascial pain. Keywords: dental occlusion, muscular oral disorder, myofascial pain, stepwise multiple regression, temporomandibular disorders, temporomandibular joint