In order to assess the quality of streams susceptible to contamination by pesticides we apply biochemical and genotoxic biomarkers in the Neotropical fish Prochilodus lineatus submitted to in situ tests. Fish were caged, for 96 h, in two streams located in areas with intensive use of pesticides, the Apertados (AP) and the Jacutinga (JC), and in a small stream (Godoy stream — GD) found inside a forest fragment adjacent to a State Park. Biochemical parameters, such as biotransformation enzymes 7-ethoxyresorufin-O-deethylase (EROD) and glutathione-S-transferase (GST), non-protein thiols (NPSH), lipoperoxidation (LPO), protein carbonylation (PCO) and acetylcholinesterase (AChE) were evaluated in various fish organs, as well as genotoxic biomarkers (damage to DNA and occurrence of micronuclei and erythrocyte nuclear abnormalities). Samples of water and sediment were collected for analysis of metals (Cu, Cr, Pb, Ni, Mn, Cd and Zn), organochloride pesticides, and triazine and glyphosate herbicides. We observed an increase in liver GST activity in fish at AP and gill GST activity in fish at JC. An increase in liver LPO was also observed in fish exposed to AP and JC. The same animals also exhibited increased DNA damage and erythrocyte nuclear abnormalities (ENAs) compared to the fish kept in GD. A number of compounds showed concentrations higher than the permitted levels, in particular, dichlorodiphenyltrichloroethane (DDT), its metabolites dichlorodiphenyldichloroethylene (DDE) and dichlorodiphenyldichloroethane (DDD), hexachlorocyclohexanes (HCH), heptachloride, diclofluanid and aldrins. These pesticides were detected at higher concentrations in water and sediment samples from AP, followed by JC and GD. The Integrated Biomarker Response Index (IBR) indicated that AP and JC (AP: 21.7 > JC: 18.5 > GD: 12.6) have the worst environmental quality. Integrated biomarker analysis revealed that the alterations observed related well with the levels of environmental contaminants, demonstrating the effectiveness of this biomonitoring approach.