Aims: To investigate whether any of the pretreatment physical signs, symptoms, and responses on psychological questionnaires would predict treatment outcomes after a standardized temporomandibular disorder (TMD) treatment program. Methods: The care provided to 157 TMD patients was a short course of physical therapy, an occlusal appliance, and over-the-counter nonsteroidal anti-inflammatory drugs (OTC NSAIDs). A multidimensional outcome assessment was performed using six variables. Follow-up data were available on 81.5% of enrolled subjects and elapsed time from initial visit to the two follow-up points was 13 ± 4.7 and 33.6 ± 9.8 months, respectively. Multiple regression analyses were conducted to assess the relationship between 18 predictor variables and the six outcome variables. Results: The results showed that the combination of a higher initial visual analog scale (VAS) pain score plus a lower jaw function interference score was significantly associated with a reduction of VAS pain after treatment (P .05; adjusted R2 = 0.54). Moreover, the combination of a higher initial activity limitation score plus a lower jaw function interference score was associated with a greater reduction of the activity limitation score after treatment (P .05; adjusted R2 = 0.36). None of the other outcomes were found to relate to any of the pretreatment variables. It must be noted that no single variable was a strong predictor and the odds ratios between the above three variables and the predicted outcomes were not robust. Conclusion: The corollary of these results suggests that if a high degree of jaw function interference is present (eg, clicking, locking), then the prognosis of improvement with brief self-directed physical therapy, an occlusal appliance, and OTC NSAID is lower, at least within the time frame of this study. Palabras clave: multiple logistic regression analysis, physical medicine treatment, predictors, temporomandibular disorders, treatment outcome