[en] To study the reliability between in vitro and in vivo data collected downstream 2 sewage treatment plants (STP) as well as from bleached kraft mill industry (BKME), 5 rivers (3 impacted and 2 references) were investigated in the Walloon region (southern of Belgium). For the in vitro part of the work, water samples were collected to measure the estrogenicity of the ‘out’ effluent compared to reference sample point by MCF-7 assay. Results indicated significant estrogenicity of effluents from STP and BKME and a weak estrogenicity in reference sites. However, estradiol equivalents (EEQ) estimated into rivers were probably too low to impact wild population. Chemical analysis of 13 compounds of interest indicated that extraction procedure used in this study gave low recoveries of estrogen-like xenobiotics, leading to probably under-estimated MCF-7 responses. Surprisingly, a full scan mode has revealed an unexpected compound in the sample of BKME which was: 7-isopropyl-1,1,4a-trimethyl-1,2,3,4a,9,10,10a-octahydrophenanthrene, a product of pulp mill manufacture. In parallel to in vitro, in vivo assessment of estrogenic impact of effluent was followed on the gudgeon (Gobio gobio). Samples were achieved during 2 different periods of the reproductive cycle, resting period (RP) and pre-spawning period (pSP). Unspecific physiological parameters to estrogenic exposure (gonadosomatic index and systematic testis cell counting) displayed no significant differences related to endocrine disruption of the reproductive tract, only differences were correlated with the reproductive state of fish (RP versus pSP). Concerning the potent biomarker of estrogen exposure, vitellogenin (vtg), only basal induction was revealed but not related to estrogenic exposure. Nevertheless, vtg over-expression was found for male fish presenting a feminization of the reproductive tract captured downstream the STP station of Wégnez in the Vesdre River. Intersexuality, another indicator of the estrogenicity impact in fish, was observed in every site. Actually, ovotestis was systematically formed by protoplasmic oocyte observed in low percentage in every group analysed (impacted and references). Moreover, in fish captured in Wégnez, oocyte diameter was significantly higher compared to the other groups. In this study, only moderate to none impact in population of gudgeon was noticed. Moreover, in this case no discrepancy between in vitro and vivo was viewed although both approaches revealed gaps in monitoring effluent incidence into the environment. We should remain careful in the interpretation when only partial approaches are used in order to characterize impact in the aquatic milieu.