Pollution of water ecosystems is considered one of the most serious problems in the current days. In order to predict the effects of pollutants present in the water before overt disease and/or death in aquatic biota, bioindicators, as fish, and biomarkers have been widely used in biomonitoring studies. Notwithstanding, biomarkers responses might be modulated by non-pollutant factors and not only by environmental contaminants. As these premises are not always observed, this review attempts to explore confounding factors related to the modulation of the biomarker responses applied in biomonitoring studies. Among others, gender, temperature and feeding habitat are some factors which affect biomarker responses. However, the use of this tool is still important to assess the health of aquatic ecosystems. Thus, it is strongly recommended to use different biomarkers and, whenever possible, to analyze water and sediment physicochemical parameters.