Diploid grass carp (Ctenopharyngodon idella L.) were introduced to a lowland Waikato drainage canal at an initial density of 40-80 kg ha -1(83-167 fish ha -1) to control aquatic macrophytes and improve water flow. A near-by canal was left without grass carp to act as an untreated control. After 7 months, macrophytes occupied 17% of the water column in the treated canal compared to 78% in the untreated canal. Fish and macroinvertebrates in both canals were examined before and after the release of grass carp by sampling with replacement by fyke netting on seven occasions. Brown bullhead catfish (Ameiurus nebulosus (Lesueur)) and shortfinned eels (Anguilla australis Richardson) comprised most of the resident fish biomass in both canals; however, before grass carp stocking, eels were more abundant than catfish in the treated canal. There was no change in the abundance of resident fish after stocking, but young-of-the-year catfish had greater mortality and grew faster in the treated canal than in the untreated canal. Macroinvertebrates were primarily associated with aquatic macrophytes. Grass carp reduced aquatic macrophyte abundance in the treated canal by about 80%, which by inference reduced the abundance of associated macroinvertebrates, but there was no observed impact of grass carp stocking on the resident fish assemblage. We examined the relationship between head width and fish length, and from this determined that 70% of the grass carp could have escaped through the downstream retention screen. Despite this possibility, grass carp remained in the canal and effectively controlled aquatic macrophytes for 18 months.