The assays developed for the detection of endocrine disruptors (ED) are often used to assess the effect of a single compound although in reality, organisms are often exposed to complex mixtures of compounds. To date, studies on ED mixtures have mainly focused on the potential additive effects of ED, especially chemicals that mimic 17-ß-estradiol responses. Until now most in vitro and in vivo assays that have been employed to address mixture effect focused on ERa and liver vitellogenin expression living unexplored mixture effect of estrogens on other ER subtypes and ER-regulated genes. In this context, the aim of our project is to determine the effects of mixtures of estrogenic compounds in fish by using zebrafish specific in vitro and in vivo bioassays allowing assessment of effects of single compounds or mixtures on the three zebrafish ER subtypes and on the brain specific ER-regulated cyp19a1b gene. For that purpose, newly established zebrafish tools are used: (1) in vitro stable reporter gene assay (ERE-luciferase) in zebrafish liver-derived cells (2) in vitro reporter gene assay (ZfCyp19a1b-luciferase) in human glial cells, both transfected with either zfER a, ß1 or ß2 and (3) in vivo cyp19a1b-GFP transgenic zebrafish. In this project we focused on binary mixtures of full ER agonists, partial ER agonists and aromatizable androgens. Concentration-response relationships for all single chemicals were measured, modeled, and used to design the binary mixture experiments following a ray design. The mixture experiments results were analyzed to predict joint effects according to concentration addition (CA) and independent action (IA) models. Two different statistical approaches (Jonker's dose-response surface models and Hewlett and Streibig's isobole-based models implemented by Sorensen) are used to assess deviations from the CA model and to characterize interactions between the components of the mixtures (synergism/antagonism). The first results of the project will be presented.