Aims: To examine the prevalence of different substances used by adolescents admitted to hospital emergency departments (ED); to evaluate the impact of an ED based brief intervention (BI) on hospital events; to compare outcomes for those using 'alcohol alone', 'alcohol plus illicit licit drugs' ('alcohol plus'), or 'other drugs' excluding alcohol, and investigate the relationship between hazardous alcohol consumption patterns and hospital events.Design: We used hospital record linkage to follow-up a randomised control trial cohort.Participants: Adolescents (12-19 years) recruited in ED with presentations involving alcohol or other drugs (AOD): 67 received usual care and 60 a BI that facilitated attendance at community drug agencies.Measurements: Drug-use categories were assigned from the substances used at the baseline presentation. Outcomes were assessed as hospital admissions plus ED presentations in the 12-month post-intervention. 'Hazardous' alcohol use was categorised via the AUDIT-3.Results: The drug-use categories were 'alcohol alone' (n = 67, 53%), 'alcohol plus' (n = 31, 24%) and 'other drugs' (n = 28, 22%). In the 12-month post-intervention, the randomisation groups had similar numbers of AOD hospital events. A Cox regression showed that in the usual care but not the BI group, for 'other drugs' there was a 8-fold increased hazard ('risk') of an AOD hospital event compared with 'alcohol alone' and a 10-fold increase compared to 'alcohol plus'. Each pre-recruitment AOD event doubled the hazard of an AOD event. For the 131 group, these were not significant predictors. The 'other drugs' group had more AOD events than either of the other groups. 'Hazardous' (77%) alcohol use was common but was not a predictor of AOD hospital events.Conclusions: : BI can be delivered in ED and reduce hospital AOD morbidity associated with the use of drugs other than alcohol. Interventions should focus on those with prior AOD events and 'other drugs' presentations. (c) 2005 Elsevier Ireland Ltd. All rights reserved.