In order to determine if a relationship exists between the clinical symptoms of TMJ dysfunction and the reproducibility of mandibular movements, twenty one subjects were chosen. The control group consisted of 5 subjects who were determined to be free from signs and symptoms of dysfunction . The sixteen experimental subjects were selected on the basis of their having dysfunctional symptoms. The author obtained two sets of pantographic tracings. Each set consisted of tracings from three both lateral and one protrusive movements. A second set of tracings were recorded immediately on the new recording papers using the same procedure as the first tracing. The tracings were scored by Pantographic reproducibility index (PRI). The obtained results were as follows. 1. Mean PRI scores of groups increased as the degree of dysfunction were increased. 2. For the groups of no or slight dysfunctional symptoms the PRI scores of the second tracing were smaller than the first one, wherea the scores of the second tracing from $D_2,;D_3$ group were larger than the first tracing. 3. Differences between the mean PRI scores of control group and those of experimental group were statistically significant. 4. The second sets of tracings were more reliable statistically than those of first ones. 5. PRI can be used as a meaningful aid for the evaluation of the diagnostic and therapeutic results of treatment modalities for the TMJ dysfunction. 6. At 3east two sets of tracings should be recorded when the PRI is to be used to detect the incoordinated movements of TMJ dysfunction patient. 7. PRI scores of control group ($D_0$) was 137.7, thus, mandibular movement was reproducible, whereas PRI scores of experimental groups ($D_1,;D_2,;D_3$) were 22.5, 27.7, 30.45 respectively, thus were nonreproducible.