The evidence that the effects of the antidepressant duloxetine on painful physical symptoms in depression and chronic pain disorders are a direct analgesic effect rather than an indirect antidepressant effect is reviewed. Data from placebo-controlled acute studies of duloxetine in major depressive disorder, diabetic peripheral neuropathic pain and fibromyalgia syndrome are included in this review. In placebo-controlled studies of duloxetine in patients with major depressive disorder, non-depressed diabetic peripheral neuropathic pain, and fibromyalgia syndrome, duloxetine has a statistically significantly greater effect on pain than placebo. Path analysis suggests that in these patient populations, approximately 50, 90, and 80%, respectively, of the observed effect on pain is a direct analgesic effect rather than an indirect antidepressant effect. In fibromyalgia syndrome studies, duloxetine had similar and substantial effects on pain regardless of whether patients had comorbid major depressive disorder. Pain is a complex experience, involving both the physiological responses of the nociceptive system and the processing of that information in brain regions associated with emotion. While some effects of duloxetine on painful symptoms can be accounted for by its antidepressant action, the data strongly suggest that duloxetine also exerts a substantial direct analgesic effect over and above its antidepressant effects, in patients with major depressive disorder, diabetic peripheral neuropathic pain, and fibromyalgia syndrome.