Across New Zealand, the institute operates 150 hydrometric stations recording continuous flow at 15-minute intervals and undertakes monthly water quality sampling at 77 sites.
Many more sites are monitored by regional councils, which use Niwa's high-quality data as benchmarks for their sites.
Proposed changes, now being consulted on, include Niwa setting up a number of new "best-practice benchmark sites" to help other groups and agencies set up and test their own operations, strategy general manager Dr Bryce Cooper said.
The changes would have no negative impact on the quality of freshwater testing, he said. It was likely the outcome of the review would result in Niwa monitoring fewer sites, "but to much higher standards".
"The aim of the current consultation is to ensure that regional council monitoring operations, which consist of about 1000 freshwater monitoring sites across the country, are consistent and provide a reliable dataset for national assessments."
Niwa would also research and share new monitoring techniques, maintain a high-quality dataset of New Zealand freshwaters and work with groups to ensure a "consistent national monitoring programme".
"Currently, national analyses of Niwa and regional council datasets are compromised by information inconsistencies and data coverage issues," Dr Cooper said.
Green Party environment spokeswoman Eugenie Sage said the National River Water Quality Network's 23 years of data had been "invaluable" in showing changes in water quality since 1999 and providing some baseline information from before the main dairying boom.
"Any cuts to this nationwide rivers programme, by reducing the number of sites or how often they are monitored, would come at the worst possible time because Government has cancelled the New Zealand-wide state of the environment report due in December," she said.