This several year study investigated the nitrogen and phosphorus removal capabilities for a full-scale wetland treatment system at a campground and conference center in Centre County, PA, The lined, subsurface-flow treatment wetlands, designed to treat 17 m(3) day(-1), were constructed in parallel and vegetated with Typha latifolia (broadleaf cattail) and Schoenoplectus tabernaemontani (softstem bulrush), Effluent from the two treatment wetlands discharged into two parallel, unlined sand filters. One sand filter was modified to distribute waste at the influent end (header-type), while waste was distributed over the surface of the second sand filter through a pipe network (surface-distributed type). The system consistently removed CBOD(5) and TSS to below 30 mg L(-1) In the first three years of operation, system effectiveness at removing the total nitrogen load increased from 60.1% to 93.4%, primarily due to increased vegetation densities in the subsurface wetland cells. Parallel wetland treatment cells had significantly different plant densities, with significantly more total N removed from the more densely vegetated cell, Over 99% of the dissolved phosphorus was removed in the three year study, but removal efficiencies within the treatment cells decreased with time, The header-type sand filter was significantly more effective at nitrifying the waste than the surface-distributed sand filter. The ability of the sand filters to nitrify waste could be improved by converting the surface-distributed filter to a header-type system. Nitrogen removal in the wetland treatment cells may continue to improve as the density of wetland vegetation increases, while phosphate removal efficiencies may decrease as soil adsorption sites become saturated with time.