Objectives To evaluate the correlations between temporomandibular joint (TMJ) pain and both conventional and gadolinium-enhanced magnetic resonance (MR) imaging characteristics in patients with temporomandibular disorders (TMDs). Methods T1-weighted, T2-weighted, and fat-suppressed contrast-enhanced T1-weighted MR imaging were performed in 80 TMD patients with the main complaint of TMJ pain. A kappa statistical analysis was performed to calculate the interobserver agreement between two of three radiologists. The relationships between TMJ pain and various MR findings were analyzed by multiple comparison test, Spearman's rank correlation test, Chi-square test, Student's t test, and multivariate logistic regression analysis. Results Significant correlations were observed between anterior disk displacement without reduction, joint effusion, and TMJ pain (p \ 0.01). The degree of contrast enhancement of the posterior disk attachment was correlated with the severity of TMJ pain, especially spontaneous pain (r = 0.725, p \ 0.01), while no correlation was observed in the masticatory muscle pain group. The correlation tended to be higher in the latero-central portions than in the medial portion of the mandibular condyle, although no significant difference was seen. Conclusions It has been suggested that fat-suppressed contrast-enhanced MR imaging techniques facilitate better understanding of the sources of TMJ pain, but not masticatory muscle pain, and could reflect TMJ synovial inflammation. The severity of TMJ pain was closely correlated with the degree of contrast enhancement. The above-mentioned correlation had a tendency to be higher in the latero-central portions of the mandibular condyle.