Treatment wetlands (TW) operated as bioelectrochemical systems (BES-TW) provide a higher degree of treatment than conventional TW. Yet, the fundamental processes or mechanisms for the envisaged better performance of BES-TW over conventional TW remains poorly understood. This work aimed to determine to which extent microbial activity enhancement could be the reason behind this treatment performance increase. To this purpose, pilot-scale horizontal sub-surface flow BES-TW operated under three different configurations were continuously fed with real urban wastewater. BES-TW were evaluated for COD and ammonia removal efficiency, and two techniques of microbial activity assessment were applied. Configurations, tested in duplicate, were: control TWs without electrodes (C-TW), TWs operated as microbial fuel cells (MFC-TW), and TWs operated as microbial electrolysis cells (MEC-TW). Microbial activity was assessed by measuring the enzymatic activity (EA) (FDA hydrolysis technique) and the aerobic activity (AA) (estimated through respirometry). Results showed that BES-TW outperformed C-TW in terms of both microbial activity enhancement and contaminants removal efficiency, especially in the case of MEC-TW. More precisely, this configuration showed an average improvement of 17%, and 56% in COD removal and EA efficiencies, respectively, compared to C-TW. Regarding AA activity, although MEC-TW seemed to outperform the rest of the configurations, differences were not statistically significant. This work demonstrates that TWs operated as BES increase the overall enzymatic activity of the treatment bed and this, in turn, is the leading cause to a higher degree of treatment performance.