The objectives of this research were to determine the biomass of koi carp (Cyprinus carpio) and other fish species in Lake Ohinewai in 2016 by two-sample mark-recapture methods and to compare these estimates with similar studies conducted in 2011, 2012, and 2014. In May 2011 the University of Waikato installed a one-way barrier on the outlet of Lake Ohinewai that was designed to prevent adult koi carp moving upstream into the lake, while allowing fish to pass downstream out of the lake, thereby attempting to passively reduce fish biomass in the lake. In addition, an invasive fish removal programme was undertaken in Lake Ohinewai with the aim of reducing the koi carp population from the estimated original biomass of 308 kg/ha (211–466, 95% CL) to below 100 kg/ha. In previous mark-recapture studies it was estimated that the koi carp biomass had been reduced to 39 kg/ha (24–67, 95% CL) in 2012 and 14 kg/ha (7–27, 95% CL) in 2014 by a combination of fish removal and the one-way gate. In 2016, we estimated that the koi carp biomass had increased to 94 kg/ha (49–197, 95% CL). A mark-recapture study was conducted on the fish community in Lake Ohinewai from 21 November to 7 December 2016. Capture method included fyke net sets for a total of 120 net nights and boat electrofishing for 12 h. Species included in the study were the invasive species koi carp, brown bullhead catfish (Ameiurus nebulosus), rudd (Scardinius erythrophthalmus), goldfish (Carassius auratus) and koi carp-goldfish hybrids, and the native species shortfin eel (Anguilla australis) and longfin eel (Anguilla dieffenbachii). In addition, water samples for nutrient analysis and suspended solids, Secchi depth, CTD (conductivity, temperature, depth) profiles and zooplankton and phytoplankton samples were taken from Lake Ohinewai on three occasions between 21 November 2016 and 20 January 2017. A total of 1680 fish were caught in the marking phase (21–24 November 2016) and 2058 fish in the recapture phase (5–8 December 2016); 479 of fish caught in the recapture phase were marked. The estimated koi carp population in Lake Ohinewai more than quadrupled in size from 454 (251–889, 95% CL) fish in 2014 to 2063 (1070–4328, 95% CL) fish in 2016. Similar increases in the estimated populations of catfish (925 to 4010), goldfish (512 to 1927) and koi carp hybrids (43 to 252) were also observed. Eels also increased in abundance; between 2014 and 2016 shortfin eels increased from 2305 to 3456 and longfin eels increased from 44 to 100. Total invasive fish biomass also increased from 29 kg/ha (18–50, 95% CL) in 2014 to 154 kg/ha (78–311, 95% CL) in 2016. This change was primarily due to an increase in koi carp biomass, but increases in catfish and goldfish biomasses have also occurred since 2014. Goldfish showed a strong pulse of recruitment in 2016. The large-scale removal of invasive fish and installation of the one-way barrier resulted in significant changes to the invasive fish community composition. The fish removal programme reduced the proportion of larger (>275 mm fork length) koi carp and goldfish, which can be attributed to size selectivity of the removal methods and emigration of adult koi carp from the lake. Recruitment of smaller koi carp was observed in 2014 when they had bimodal size distribution, comprising juvenile fish (<250 mm FL) and adult fish (>300 mm FL). Shortfin eels increased significantly in mean weight following carp removal. Changes in the longfin eel population appeared similar to those of shortfin eels, with an increased proportion of larger eels following removal of koi carp and other invasive fish, but the low sample size hampered comparisons of mean weight. Water quality indicators such as nutrient concentrations, Secchi depth and total suspended solids showed no improvement in the 5 years following the reduction in invasive fish biomass, except for chlorophyll a concentration, which declined with decreasing biomass of invasive fish from 2011 to 2014, and returned to previous concentrations as invasive fish biomass increased by 2016. It is not entirely clear that this was caused by changes in the fish biomass as total nitrogen also declined from 2009 to 2013. Total phosphorus remained largely unchanged throughout the study period. In addition, phytoplankton species and abundance appear consistent with other hypereutrophic lakes in the Waikato region. The annual mean TLI3, which excludes Secchi depth, ranged from 6.1 to 6.6 between 2006 and 2017, indicating that the lake remained hypertrophic during fish removal. Chemically-driven internal nutrient cycling and catchment nutrient inputs are likely to maintain the hypereutrophic condition and low water clarity will continue to inhibit re-establishment of submerged macrophytes, the absence of which can exacerbate wind resuspension of sediment. The increase in the koi carp abundance indicates that at least one successful recruitment event occurred following the removal programme and installation of the one-way barrier, either by spawning in the lake or by upstream migration of juvenile young-of-the-year fish from spawning areas lower in the catchment. In addition, reduced interspecific competition has likely contributed to increased recruitment of the catfish and goldfish populations. Therefore, while the Lake Ohinewai one-way barrier appears effective in preventing immigration of adult koi carp into an area, it should not be considered as a mechanism to exclude both adult and juvenile koi carp nor as a sole measure of control. In Lake Ohinewai, reductions in koi carp biomass were not long lasting, and during the period of low carp abundance there was no convincing evidence for a corresponding improvement in water quality and zooplankton and phytoplankton community composition. Koi carp removal did appear to increase the abundance and size of shortfin eels. Before any restoration programme is initiated it is recommended that a complete assessment of the system is undertaken to determine which factors are driving the decline in water quality. Ecological models can be helpful in determining these driving factors but they are reliant on substantial data for parameterisation and calibration. If reductions in invasive fish biomass are deemed necessary for ecosystem restoration, repeated removal programmes appear to be the only viable way of ensuring low biomasses of target invasive species are maintained