Objectives This study was performed to assess the relationship between plasma levels of milnacipran and its analgesic/antidepressive effect in patients with chronic orofacial pain treated with this drug. Methods A total of 44 patients took milnacipran for 12?weeks. Patients were assessed for their pain and depressive symptoms using the visual analog scale (VAS) and Hamilton Depression Rating Scale, respectively. The plasma milnacipran level was also assessed at week 12. Results Forty patients completed study treatment and were included in the analysis. In these patients, the VAS score at week 12 significantly decreased from the baseline score (t?=?5.15, p?<?0.0001). The dose of milnacipran was positively correlated in a linear manner with the plasma level of the drug (Y?=?44.86?+?0.33X, r?=?0.54, R2?=?0.29, p?=?0.0004). A quadratic regression curve was plotted between the percentage of decrease in the VAS score and plasma milnacipran level (Y?=?27.39?+?0.76X?-?0.008X2, p?=?0.048, r?=?0.40, R2?=?0.16). On the other hand, no significant relationship was noted between the percentage of decrease in the Hamilton Depression Rating Scale score and plasma milnacipran level. Conclusion The analgesic effect of milnacipran was suppressed in the presence of the plasma level of the drug outside the therapeutic range, whereas its antidepressant effect was not affected by its plasma level. Copyright (c) 2012 John Wiley & Sons, Ltd.