Bioelectrochemical system (BES) based biosensors for organic sensing has long been investigated. However, there is no uniform criterion to evaluate directly the performance of the BES based biosensors due to their different scale. Here, for the first time, we show that the normalized maximum detection range (NMDR) and normalized sensing time (NST) can potentially be used as the two criteria in BES based biosensors for organic sensing. Thereafter, the recently emerged, relatively larger scale BES (i.e. constructed wetland-microbial fuel cell, CW-MFC) was specifically examined in this study. The biocathode formation and the influence of anodic material on sensor performance were systematically evaluated. The system with metal-based anode was found to produce a more stable and quicker response (low NST) than that with carbon-based anode. Significantly, the continuous loading mode was found to greatly reduce the NMDR compared to the batch mode, and the hydraulic residence time (HRT) is the critical factor determining the NMDR. Furthermore, it was found that the electrical signals generated from the CW-MFC system were insignificantly influenced by some specific chemical disturbances, such as Cu2+ and herbicide. Therefore, normalized toxicity (NT) is suggested to be considered in BES based biosensor. However, for chemicals with higher reduction potentials (NO3- in this work), the system presented a high response, enabling its potential for monitoring NO3- in effluents or groundwater. This study can hopefully contribute to further development of the sustainable BES based biosensors in CW.