Introduction. Data from the 2005 Health Barometer on perceptions, attitudes, and behaviors regarding the use of psychoactive substances allow for analyses based on educational attainment, income, and employment status. Analyzing responses according to these criteria provides an opportunity to highlight social inequalities, enabling their consideration in the implementation of prevention and health promotion initiatives. Materials and Methods. The 2005 Health Barometer, a cross-sectional survey representative of the French population living in ordinary households, was conducted by telephone on a random sample of 30,514 individuals aged 12 to 75 (overall refusal rate: 42.1%). Thanks to a specific procedure for generating telephone numbers, households on the do-not-call list, as well as those equipped solely with a mobile phone (12.5%), were also included in the sample. Only one person per household participated in the survey. As part of this study of 16- to 64-year-olds, logistic regressions controlling for key structural effects were applied to indicators of psychoactive substance use. Results. While some nuances emerge depending on whether the substance in question is tobacco, alcohol, cannabis, or other drugs, for all observed usage behaviors, high socioeconomic and educational status appears associated with experimentation with these substances but not with regular use, or even use harmful to health. Conversely, unemployment is linked to such practices, even though it is less associated with experimentation. Discussion and conclusion. The 2005 Health Barometer highlights differences in reported psychoactive substance use depending on the sociodemographic variables considered. Although the links between health and work are far from straightforward, unemployment appears particularly linked to reports of regular use, and even situations of dependence. Overall, it seems that people with lower levels of education experiment less with psychoactive substances but report more frequently being in situations of more problematic use. Educational attainment likely facilitates a better understanding of prevention messages. Through its methodology, the Health Barometer provides valuable cross-sectional data for observing health inequalities. However, given the limited number of indicators available, the analyses that can be conducted remain fragile in terms of their interpretation. Future Health Barometers will include more specific questions on various situations of vulnerability, incorporating dimensions such as the forgoing of medical care, for example. [author’s abstract] Paper cited in the Revue d’Épidémiologie et de Santé Publique, volume 56, issue 6.