Four wastewater samples of different treatment qualities; untreated, alarm, alert and normal, from a Swedish chemi-thermo-mechanical pulp mill and pure water were investigated using an amperometric bio-electronic tongue in a batch cell. The aim was to explore enzymatically modified screen-printed amperometric sensors for the discrimination of wastewater quality and to counteract the inherent drift. Seven out of eight platinum electrodes on the array were modified with four different enzymes; tyrosinase, horseradish peroxidase, acetyl cholinesterase and butyryl cholinesterase. At a constant potential the current intensity on each sensor was measured for 200s, 100s before injection and 100s after injection of the sample. The dynamic biosensor response curves from the eight sensors were used for principal component analysis (PCA). A simple baseline and sensitivity correction equivalent to multiplicative drift correction (MDC), using steady state intensities of reference sample (catechol) recordings, was employed. A clear pattern emerged in perfect agreement with prior knowledge of the samples explaining 97% of the variation in the data by two principal components (PCs). The first PC described the treatment quality of the samples and the second PC described the difference between treated and untreated samples. Horseradish peroxidase and pure platinum sensors were found to be the determinant sensors, while the rest did not contribute much to the discrimination. The wastewater samples were characterized by the chemical oxygen demand (COD), biological oxygen demand (BOD), total organic carbon (TOC), inhibition of nitrification, inhibition of respiration and toxicity towards Vibrio fischeri using Microtox, the freshwater alga Pseudokirchneriella subcapita and the freshwater crustacean Daphnia magna.