Objective: To examine the demographic, prescription, ingestion, and psychiatric diagnostic factors that distinguished elderly from nonelderly patients treated for deliberate self-poisoning (DSP). Method: A prospective case series study of 2,667 patients presenting to a regional referral center for poisoning (Newcastle Mater Hospital, NSW, Australia), January 1991 to July 1998. The sample was stratified into two groups, 65 years or greater (n = 110) and 64 years or less (n = 2,557) at the time of index admission. The groups were compared using a forward stepwise logistic regression model. Uncontrolled comparisons were analyzed by chi-square statistic with Bonferroni-adjusted p values and controlled comparisons by odds ratio (OR) with 95% confidence interval (CI). Results: The elderly group represented 4.1% of the total. The logistic regression analysis found the elderly DSP group was more likely to have a longer length of stay (OR 5.90, CI 3.87–9.00), to have been prescribed “other” drugs (neither benzodiazepines, mood treatment drugs, nor paracetamol) before admission (OR 5.32, CI 3.34–8.48), to have been prescribed benzodiazepines (OR 3.15, CI 2.03–4.89), and to be diagnosed with major depression (OR 2.17, CI 1.41–3.36) than the younger group. The elderly group was less likely to have ingested paracetamol (OR 0.28, CI 0.14–0.54) or “other” drugs (neither benzodiazepines nor mood treatment drugs) in the DSP episode (OR 0.33, CI 0.20–0.54). Discussion: Elderly DSP patients differ in several important respects from younger patients. They have higher morbidity as a result of the DSP. Major depression plays a more important role. The strong relationship between benzodiazepine prescription and DSP in the elderly raises questions and possible prevention strategies.