BACKGROUND: Major depression is relatively common among patients with epilepsy, but there is currently no easy-to-use screening tool that can pick up symptoms of major depression while reliably excluding other factors such as adverse effects of antiepileptic drugs or memory problems due to temporal lobe epilepsy. OBJECTIVE: To develop and validate a brief and accurate screening instrument for major depression in epilepsy. DESIGN AND INTERVENTION: This study included a screen-development phase and a screen-validation phase. During the development phase, investigators at five outpatient epilepsy clinics in the US compiled a list of 46 symptoms that could help to identify major depression in patients with epilepsy. This preliminary screen was applied to 205 adults with epilepsy who were treated at a participating outpatient clinic; 35 of these patients had major depression according to the Mini International Neuropsychiatric Interview (MINI). Using discriminant function analysis, the investigators narrowed the preliminary list of symptoms down to the most essential items that correctly identified patients who had MINI-diagnosed major depression. The resultant screen was termed the Neurological Disorders Depression Inventory for Epilepsy (NDDI-E). The investigators used logistic regression analysis to determine whether adverse effects of antiepileptic drugs had a confounding influence on the assessment of depression with the NDDI-E. In addition, the NDDI-E was validated in an independent cohort of 229 adults with epilepsy; 71 of these patients fulfilled the MINI criteria for major depression. OUTCOME MEASURES: The outcome measures were the internal consistency and test-retest reliability of the NDDI-E, and its sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for identifying major depression in patients with epilepsy. RESULTS: The NDDI-E consisted of the following items: "everything is a struggle", "nothing I do is right", "feel guilty", "I'd be better off dead", "frustrated" and "difficulty finding pleasure". Each item was rated on a five-point Likert scale. The screen had sufficient internal consistency (Cronbach's alpha coefficient 0.85) and test-retest reliability (Spearman correlation coefficient 0.78 for two NDDI-E assessments done 2 weeks apart). At a cutoff score >15, the NDDI-E had a sensitivity of 81%, a specificity of 90%, a PPV of 62% and an NPV of 96% for predicting MINI-diagnosed major depression. A patient's adverse-event profile did not influence the screening result. Validation of the NDDI-E in the independent sample of epilepsy patients confirmed that optimum results are obtained at an NDDI-E cutoff score >15, with a sensitivity of 73%, a specificity of 72%, a PPV of 53% and an NPV of 86% for predicting major depression. CONCLUSION: The NDDI-E can help to identify symptoms of major depression in epilepsy patients; these symptoms can be differentiated from adverse events of antiepileptic treatment.