Home reverse osmosis units carry a rejection figure, and it is the number most buyers read first. It is also the number most easily misread. A rejection rate describes how much of a dissolved solid a membrane kept out under one set of conditions, and those conditions are rarely printed beside the headline percentage.
What the percentage measures
A reverse osmosis membrane separates water from dissolved salts by forcing feed water across a semipermeable surface under pressure. Clean water passes through as permeate, while most dissolved ions are carried off in a concentrate stream. Rejection is calculated by comparing total dissolved solids in the feed with those in the permeate. As a hypothetical, a feed reading of 400 and a permeate reading of 20 would correspond to a rejection of 95 percent. The arithmetic is simple. The conditions behind it are not.
The same membrane can report different rejection when conditions change. Higher feed concentration tends to lower apparent rejection. Temperature affects how readily water passes through the membrane, and warmer feed generally increases flow, which can lower the rejection reading. Operating pressure matters as well. Recovery, the share of feed water converted to permeate, is a major variable. Pushing recovery higher concentrates salts and scale-forming minerals at the membrane surface, which changes both performance and maintenance needs.
Pretreatment decides how long the number holds
Chlorine oxidizes the thin polyamide layer used in many membranes, so a carbon stage ahead of the membrane is commonly used to remove free chlorine. Sediment and iron can foul the surface and reduce flow. Hardness and silica can scale the membrane at higher recovery unless the system uses an antiscalant or runs at lower recovery. None of these problems appears in a single test reading, but each one can move the performance figure over months. A unit tested on new membranes with well-conditioned feed water will not necessarily perform the same way years later with unchanged plumbing.
Feed water is also not fixed. Changes at the source, in treatment, or in distribution can alter what enters a home, and a unit tested under one water chemistry may face another. Funding pressure adds uncertainty to that picture. Circle of Blue, in reporting by Abigail Mihaly, describes a "federal water infrastructure funding cliff" facing communities across the country. Whether any particular water system will change its supply or treatment is not something this article can determine. The practical point is that the water a household's unit receives may shift without any change to the unit itself.
What to ask before comparing figures
Ask the seller for the feed total dissolved solids at the time of testing, along with operating pressure, feed temperature, recovery rate, membrane age, and the sampling point. Ask whether the rejection figure comes from certification to a recognized standard, and which contaminants that certification covers. Ask whether the figure applies to total dissolved solids generally or to a specific contaminant. Many figures in sales material are not directly comparable, and a higher percentage printed on one sheet may rest on gentler test conditions than a lower one on another.
A handheld total dissolved solids meter can read feed and permeate on the same day and give a rough rejection estimate. That is a screening check, not a laboratory analysis. It does not confirm removal of specific regulated contaminants such as lead, nitrate, or PFAS, which require targeted testing. Keep the readings, the date, and the sampling location together so that a later comparison has the conditions attached.
Households that want a fuller picture can have local RO specialists test the unit against the actual feed water and set a schedule for prefilter and membrane service. The record of those tests matters as much as the result, because the rejection figure is only useful when it can be tied to the water and conditions that produced it.