Conventional models for predicting the fate of xenobiotic organic trace chemicals, identified, and calibrated using data obtained in batch experiments spiked with reference substances, can be limited in predicting xenobiotic removal in wastewater treatment plants (WWTPs). At stake is the level of model complexity required to adequately describe a general theory of xenobiotic removal in WWTPs. In this article, we assess the factors that influence the removal of diclofenac and carbamazepine in activated sludge, and evaluate the complexity required for the model to effectively predict their removal. The results are generalized to previously published cases. Batch experimental results, obtained under anoxic and aerobic conditions, were used to identify extensions to, and to estimate parameter values of the activated sludge modeling framework for Xenobiotic trace chemicals (ASM-X). Measurement and simulation results obtained in the batch experiments, spiked with the diclofenac and carbamazepine content of preclarified municipal wastewater shows comparably high biotransformation rates in the presence of growth substrates. Forward dynamic simulations were performed using full-scale data obtained from Bekkelaget WWTP (Oslo, Norway) to evaluate the model and to estimate the level of re-transformable xenobiotics present in the influent. The results obtained in this study demonstrate that xenobiotic loading conditions can significantly influence the removal capacity of WWTPs. We show that the trace chemical retransformation in upstream sewer pipes can introduce considerable error in assessing the removal efficiency of a WWTP, based only on parent compound concentration measurements. The combination of our data with those from the literature shows that solids retention time (SRT) can enhance the biotransformation of diclofenac, which was not the case for carbamazepine. Model approximation of the xenobiotic concentration, detected in the solid phase, suggest that between approximately 1% and 16% of the total solid carbamazepine and diclofenac concentrations, respectively, is due to sorptionthe remainder being non-bioavailable and sequestered. We demonstrate the effectiveness of the model's predictive power over conventional tools in a statistical analysis, performed at four levels of structural complexity. To assess WWTP retrofitting needs to remove xenobiotic trace chemicals, we suggest using mechanistic models, e.g., ASM-X, in regional risk assessments. For preliminary evaluations, we present operating charts that can be used to estimate average xenobiotic removal rates in WWTPs as a function of SRT and the xenobiotics mass loads normalised to design treatment capacity. Biotechnol. Bioeng. 2012; 109: 27572769. (c) 2012 Wiley Periodicals, Inc.