While there has been substantial community discussion and concern expressed about volatile substance use (VSU), there has been little research on the use and related harms of these substances compared to other drugs. In this study we address a need in existing epidemiological research on VSU harms by describing the incidence and characteristics of VSU ambulance attendances between August 1998 and May 2004 across metropolitan Melbourne relative to heroin attendances, a drug class that has received more research attention. Our analysis showed that the crude rate of VSU attendance (5.03 per 100,000 population) over the period was substantially lower than the rates of heroin "involved" and heroin "overdose" attendances (33.40 and 54.87 per 100,000, respectively). Mean age of VSU cases was 20, with users on average 8 years younger than heroin cases. Two-thirds of VSU cases were male, with the likelihood of male attendance similar to heroin involved, but significantly less likely than heroin overdose. VSU attendances were geographically more evenly distributed than heroin attendances, with VSU cases more likely to occur at public and outdoor spaces. VSU cases were also less likely to be in an altered conscious state than heroin cases, but more likely to be co-attended by police and accept transportation to hospital. We conclude that VSU and heroin related harms occurred in different cohorts across metropolitan Melbourne, and that ambulance data can supplement existing data sources to inform policy and programme development, and the monitoring of VSU harms.