A study was carried out in Zimbabwe to evaluate the performance of duckweed ponds as an option for treating and reusing wastewater in small, decentralised communities. The study focused on nitrogen and phosphorus removal, operational problems, and duckweed application. Two full-scale trial plants at Nemanwa and Gutu-Mupandawana growth points were used. Sewage samples were collected and analysed monthly from September 2000 to August 2001 for NO3, NH4(-)1 TKN, TP, COD, and other field measurements. The duckweed was harvested daily and fed to chickens. The Nemanwa plant had high nutrient levels due to nil outflows caused by water rationing in the area. The Gutu effluent had averages of 38.7 +/- 23.1 mg/l TN and 7.5 +/- 2.4 mg/l TP which are above the respective Zimbabwean standards of 10 mg/l TN and 1 mg/l TP. COD removal efficiency at Gutu was poor at 45%. The performance of Gutu and Nemanwa plants suffered from inappropriate design especially pond depth and short-circuiting. The duckweed died off in the November-January period, this being attributed to excessive levels of ammonia. It was concluded that the duckweed pond systems would offer a good alternative for managing and reusing wastewater at community level provided due regard is paid to appropriate design criteria.