Background: Duloxetine is a selective serotonin and norepinephrine reuptake inhibitor approved for the treatment of diabetic peripheral neuropathic pain (DPNP). The current analyses aimed to identify and evaluate the effect of any significant covariates on DPNP treatment response, via the development of a continuous descriptive Pharmacokinetics/Pharmacodynamics (PK/PD) model for pain score reduction and a proportional odds PK/PD model describing the proportion of patients achieving pain relief. Methods: A total of 1139 patients received placebo, 20, 60 or 120 mg duloxetine daily for 12 weeks. The primary efficacy measure was 24-h pain scores collected on the 11-point categorical numerical rating scale averaged over a week. PK/PD models were fitted using non-linear mixed-effects models. Results: Baseline pain severity was found to be an important factor in both PK/PD models. Larger drops in pain scores were observed for patients with more severe pain. The proportional odds PK/PD model used an a priori definition for adequate pain relief, which was a decrease in two points from baseline. Simulations showed that approximately 70% of patients in the highest dose groups would obtain pain relief at week 12, although placebo response was relatively high at 40%. The proportion of patients who obtained pain relief was slightly lower in those with mild pain compared to those with more severe pain. Conclusions: Patients with more severe pain at study entry had larger treatment responses and were more likely to achieve clinically meaningful pain relief with similar amounts of drug, compared to patients with milder pain. © 2013 European Federation of International Association for the Study of Pain Chapters.