OBJECTIVE: To examine the feasibility of establishing a database on non-fatal opioid overdose in order to examine patterns and characteristics of these overdoses across Australia. METHODS: Unit record data on opioid overdose attended by ambulances were obtained from ambulance services in the five mainland States of Australia for available periods, along with information on case definition and opioid overdose management within these jurisdictions. Variables common across States were examined including the age and sex of cases attended, the time of day and day of week of the attendance, and the transportation outcome (whether the victim was left at the scene or transported to hospital). RESULTS: The monthly rate of non-fatal opioid overdose attended by ambulance was generally highest in Victoria (Melbourne) followed by NSW, with the rates substantially lower in the remaining States over the period January 1999 to February 2001. Non-fatal opioid overdose victims were most likely to be male in all States, with the proportion of males highest in Victoria (77%), and were aged around 28 years with ages lowest in Western Australia (m=26) and highest in NSW (m=30). Most of the attendances occurred in the afternoon/early evening and towards the later days of the working week in all States. The rates of transportation varied according to ambulance service practice across the States with around 94% of cases transported in Western Australia and around 18% and 29% of cases transported in Melbourne and NSW respectively. CONCLUSIONS: It is feasible to establish a database of comparable data on non-fatal opioid overdoses attended by ambulances in Australia. This compilation represents a useful adjunct to existing surveillance systems on heroin (and other opioid) use and related harms.