The ability to predict rates and products of microbial biotransformation for a broad variety of chemical contaminants accurately is essential not only for chemical risk management but also in the context of contaminated site remediation or the development of green chemical alternatives. Existing prediction methods, however, fall short of fulfilling these needs mostly because they base predictions on chemical structure only, disregarding the microbial communities responsible for degradation and their actual metabolic potential as shaped by environmental conditions. The long-term goals of the proposed research are to develop the scientific basis and appropriate modeling algorithms for considering the metabolic potential of environmental microbial communities (i.e., the available pools of catalytic enzymes) in biotransformation prediction. It is proposed that enzyme-catalyzed biotransformation reactions are established as the explicit core elements of biotransformation prediction. The reactions so defined will serve as mechanistic basis to (i) experimentally explore the linkage between microbial community gene expression profiles and their observed potential for contaminant biotransformation, and (ii) use chemometrics and pattern analysis in high-dimensional space to mine environment-specific chemical biotransformation data for probabilities of biotransformation reactions. The resulting novel algorithms for the environment-specific prediction of biotransformation rates and products will be implemented into an existing, publically-accessible biotransformation prediction system (http://www.umbbd.ethz.ch/predict). The proposed research is highly interdisciplinary and will profit from the most recent technological and scientific advances in the fields of analytical chemistry, molecular biology and chemo-/bioinformatics to develop a ground-breaking approach for profiling the capacity of microbial communities for contaminant biotransformation.