OBJECTIVES: To determine the cost-utility of pregabalin (PRE) vs. duloxetine (DUL) over a 6 month time horizon from the perspective of US third-party payers using a decision tree analysis. METHODS: Literature searches identified clinical trials and real-world studies reporting the efficacy, tolerability, safety, adherence, opioid usage, and healthcare utilization and costs of PRE and DUL. The proportions of patients reported in the included studies were used to determine probabilities in the decision tree model. Average wholesale price was used to determine the costs of medications. The costs associated with healthcare utilization were determined from observational studies and all costs were adjusted to 2011 US dollars. Utility values for moderate to severe and mild pain were determined from EuroQoL scores. The overall utility values were determined by multiplying the utility values by the disutility values associated with adverse events. The base-case model included the FDA approved doses of PRE (300 mg/day) and DUL (60 mg/day) while “real-world” sensitivity analyses explored the effects over a range of doses (PRE 75-600 mg/day, DUL 20-120 mg/day). Monte Carlo simulations with 1000 repetitions were used to perform probabilistic sensitivity analyses (PSA) to examine uncertainty of the estimates used in the model. Outcomes from the model were expressed as cost per quality adjusted life-year (QALY). RESULTS: In the base-case model DUL cost less and was more effective than PRE (incremental cost -$187, incremental effectiveness 0.011 QALYs). Results from the real-world sensitivity analyses indicated that DUL was $16,000 and $20,667 more per QALY than PRE. Cost-effectiveness acceptability curves showed that DUL has a higher probability of being cost-effective, except at very low willingness-to-pay thresholds. CONCLUSIONS: Using a decision tree model that incorporated both clinical trial and real-world data, DUL was a more cost-effective option than PRE in the treatment of PDN from the perspective of third-party payers.