OBJECTIVES: Examine the real-world role of duloxetine vs. other treatment for low back pain (LBP). METHODS: 500 patients, ages 18-64 years, with >1 LBP diagnosis, as specified by HEDIS, and >1 duloxetine prescription within a year after LBP diagnosis were identified from a large, privately-insured claims database (>5 million lives; 2004-2006). Patients had continuous eligibility >6 months after their index duloxetine prescription (study period) and >6 months before their index prescription (baseline period) and no other duloxetine prescription during the baseline period. Duloxetine-treated patients were matched to 500 LBP patients (controls) who initiated another pharmacological or non-invasive LBP treatment in the same month from LBP diagnosis as duloxetine was initiated in duloxetine-treated patients. Controls were also matched using baseline propensity scores of duloxetine treatment, specific comorbidities, LBP diagnostic categories and baseline LBP treatments. Patients with back surgery prior to index date were excluded. McNemar tests were used to compare study period LBP treatment rates. Wilcoxon signed-rank tests were used to compare study period direct (medical and drug) costs from third-party payer perspective. RESULTS: During the 6-month study period, duloxetine-treated patients vs. controls had significantly lower rates of other pharmacological therapy (34.2% vs. 43.4% narcotic opioids, p=0.004; 30.4% vs. 43.2% NSAIDs, p<0.001; 17.4% vs. 26.2% muscle relaxants, p=0.001; 10.6% vs. 20.4% corticosteroids, p<0.001) and non-invasive therapy (16.4% vs. 35.6% chiropractic therapy, p<0.001; 13.0% vs. 34.2% physical therapy, p<0.001). Duloxetine-treated patients were also significantly less likely to have a back surgery during the study period compared with controls (0.4% vs. 2.0%, respectively; p=0.021). Average 6-month direct costs were not significantly different between duloxetine-treated patients and controls ($3554 vs. $3637, respectively). CONCLUSIONS: Duloxetine treatment in LBP patients vs. other non-surgical treatment was associated with a lower surgery rate as well as reduced rates of other non-surgical therapies without significant differences in direct costs.