Dispersant use is a controversial oil spill response technique in coastal areas. Using an experimental approach, this study evaluated the toxicity of dispersant use upon juveniles of golden grey mullet (Liza aurata). Fish were exposed for 48 hours to either dispersant only, chemically dispersed oil, mechanically dispersed oil, the water-soluble fraction of oil or to control conditions. Following exposure and a depuration period, biomarkers were assessed in fish hearts, namely the total glutathione content and the activity of four enzymes (glutathione S-transferase, superoxide dismutase, catalase and glutathione peroxides). Comparing biomarker responses between the different treatments, this study revealed that 48 hours exposure to dispersed oil (whether mechanically or chemically dispersed) resulted in a toxicity that was still detectable after a 14 days depuration period. Comparing biomarkers responses after an exposure to chemically and mechanically dispersed oil, this study suggests that chemical dispersion of the oil slick would not be more toxic than its natural dispersion under certain turbulent meteorological conditions (e.g. waves). Furthermore, the results indicated that the heart could be a target organ of interest in further studies investigating the toxicity of hydrocarbons. This study, which has been integrated into the DISCOBIOL project (Dispersant et techniques de lutte en milieu côtier: effets biologiques et apport à la réglementation), presents information of interest when attempting to provide a framework for dispersant applications in coastal areas.