The problem considered is whether self-reported substance use can be used in the estimation of recidivism risk among youths placed in secure care. The Secure Care Psychosocial Screening (SECAPS) and offending records of 447 youths admitted to detention centres in South Australia were examined. The target outcome was any new offending within 6 months of release. Use of a psychoactive substance at the time of committing the most recent offence was not a significant predictor of subsequent offending, nor was acknowledging having a problem with drug or alcohol use. In relation to the recent use of alcohol, marijuana, hallucinogens, sedatives/hypnotics, narcotics, stimulants and inhalants, only the use of alcohol and inhalants appeared to have significant relationships with recidivism. While the relationships were too small to permit using these items on their own to estimate re-offending risk, recent alcohol and inhalant use could be included as part of a broader recidivism risk assessment. [Putniņs AL. Substance use and the prediction of young offender recidivism.