PURPOSE The purpose of this study was to estimate the reliability and provide evidence for the validity of the Spanish version of the Substance Dependence Severity Scale (SDSS). METHODS A total of 211 substance abusers recruited at a public center that provides treatment for substance-related disorders in an outpatient setting were assessed. Reliability was estimated by Cronbach's α and test-retest. Validity evidence was studied by analyzing the relationships with the European Addiction Severity Index (EuropASI) and Health-Related Quality of Life for Drug Abusers (HRQoLDA). RESULTS Adequate reliability coefficients were found for the dependence scales of all addressed substances (α = 0.737 to 0.877; test-retest r = 0.796-0.952). Low internal consistency was found for the abuse scales (α = 0.329-0.694), and adequate test-retest coefficients on alcohol, cocaine and heroin (test-retest r = 0.708-0.902). The reliability of the cannabis abuse scales was inadequate. The SDSS scores showed significant relationships with the EuropASI and HRQoLDA dimensions. CONCLUSIONS The psychometric analyses validate the use of the severity of dependence scales. The SDSS abuse scales must be used taking the limitations detected into consideration.