OBJECTIVES: To examine the effect of medication choice between duloxetine and pregabalin on medication compliance outcomes among patients with diabetic peripheral neuropathic pain (DPNP). METHODS: A retrospective cohort study design was used with a large US national commercial health care claims database over 2005-2007. Patients aged 18-64 who dispensed duloxetine or pregabalin in 2006 were selected, with the first dispense date as the "index date." All individuals included were diagnosed with DPNP in the 12-month pre-index period, and continuously enrolled in both the 12 months pre- and post-index periods. Duloxetine and pregabalin cohorts were constructed based on the initial agent. Propensity score analysis was used to control for cross-cohort differences in demographics, pre-index clinical and economic characteristics, and pre-index treatment patterns. Medication compliance outcomes were examined between cohorts via medication possession ratio (MPR) and proportion of patients with MPR≥80%. RESULTS: Both the duloxetine (n = 603) and pregabalin (n=1,751) cohorts had the mean age around 55 years (54.9 vs. 55.6). Many duloxetine and pregabalin patients had cardiovascular disease (86.9% vs. 89.0%), neuropathic pain other than DPNP (80.3% vs. 83.2%), hypertension (78.8% vs. 82.6%), osteoarthritis (52.2% vs. 53.1%), and used anticonvulsants (36.3% vs. 35.6%) and opioids (32.2% vs. 28.4%). Controlling for demographics, pre-index clinical and economic characteristics, and prior medication history, duloxetine patients had significantly higher MPR than pregabalin patients (75.8% vs. 52.9%, p<0.05). The proportion of patients with MPR≥80% was also significantly higher among patients in the duloxetine cohort (47.4% vs. 27.6%, p<0.05). CONCLUSIONS: In a real world setting, medication compliance measured by MPR or proportion of patients with MPR≥80% was statistically significantly higher among DPNP patients treated with duloxetine than those on pregabalin. The results suggest that medication choices between duloxetine and pregabalin had statistically significant effects on medication compliance outcomes.