Introduction: This study tested the hypothesis that baseline ratings of fatigue/tiredness would be negatively associated with the efficacy of duloxetine on measures of pain and functional ability in patients with fibromyalgia.Methods: A post hoc analysis of pooled data from 4 double-blind, placebo-controlled studies of duloxetine in fibromyalgia was performed. The fibromyalgia impact questionnaire (FIQ) tiredness item score (0 to 10 scale) was used to define tiredness subgroups. Patients were stratified into 3 subgroups: mild (0 to 3), moderate (4 to 6), and severe (7 to 10) tiredness. Analysis of covariance models and logistic regressions were used to test treatment-by-tiredness subgroup interactions.Results: Data from the first 3 months are included in this post hoc analysis (duloxetine N = 797, placebo N = 535). At baseline, the distribution of tiredness severity in the duloxetine and placebo groups respectively was 3.64% and 3.75% mild, 16.71% and 15.57% moderate, and 79.65% and 80.68% severe. Rates of clinically significant (≥30% and ≥50%) improvement in brief pain inventory (BPI) average pain were similar across the tiredness subgroups. Tiredness severity at baseline was not negatively associated with the effects of duloxetine on patients' reports of functional ability using the FIQ total score, FIQ measures of physical impairment, interference with work, pain, stiffness, and depression and the medical outcomes study short form-36 (SF-36).Conclusions: Studies of duloxetine in fibromyalgia have demonstrated clinically significant improvements in pain and functional ability (FIQ, SF-36). This post hoc analysis of data shows that the efficacy of duloxetine among patients with fibromyalgia does not vary as a function of baseline ratings of fatigue/tiredness. © 2010 Bradley et al.; licensee BioMed Central Ltd.