Carwashes are highly water-consuming processes that require wastewater treatment before discharge into a sewer system due to the complex composition of leachate. Anionic surfactants (AS) are the main constituents of this wastewater because of their cleaning and solubilization properties; they can be potentially dangerous for the environment if not adequately treated. Constructed wetlands (CWs) are low-cost systems increasingly used to treat different types of wastewater; however, there are few studies on their use for the treatment of carwash wastewater. In this study, an innovative constructed wetland arranged in a "cascade" to simulate a wall system (WCCW) was experimented in 2010 and 2011 to treat AS. Three plant species were tested at different AS inlet concentrations (10, 50, and 100 mg L(-1)) with two hydraulic retention times (HRTs; 3 and 6 days): ribbon grass (Typhoides arundinacea (L.) Moench (syn. Phalaris arundinacea L.) var. picta; Ta), water mint (Mentha aquatica L.; Ma), and divided sedge (Carex divisa Hudson; Cd). All plant species grew constantly over the experimental period, showing a capacity to tolerate even the highest AS concentration. Using the HRT of 6 days, raising the inlet concentration increased the AS outlet concentration, with similar values for the treatments (median values of 0.13-0.15, 0.47-0.78, and 1.19-1.46 mg L(-1) at inlet concentrations in the order 10, 50, and 100 mg L(-1)). The shorter HRT led to significant differences among treatments in the reduction of outlet concentration, the best result being given by the tanks vegetated with Ma (A = 97.7 % with outlet concentration 0.35 mg L(-1)). After treatments of the WCCW, the AS content was reduced almost completely, with removal in the ranges 0.07-10.2 g m(-2) day(-1) for tanks planted with Ta, 0.10-9.1 g m(-2) day(-1) for Ma tanks, and 0.11-9.5 g m(-2) day(-1) for Cd tanks depending on the inlet concentration.