Test the specific claim the installed equipment is meant to meet. Pressure, hardness, disinfectant residual and dissolved-solids readings are not interchangeable. A result can be useful for maintenance without proving drinking-water safety.
Choose a measurement that answers the question
| Question | Appropriate evidence to discuss with the maker or laboratory | Common wrong inference |
|---|---|---|
| Has the filter become restrictive? | Inlet/outlet pressure at comparable flow | Low pressure proves contaminant breakthrough |
| Is a softener reducing hardness? | Suitable hardness measurement at defined raw and treated points | Lower TDS is required for softening |
| Is a carbon system meeting its disinfectant claim? | Test method matched to the named disinfectant, range and performance requirement | Any chlorine strip proves every carbon claim |
| Is RO performing as expected? | Manufacturer-specified operating checks, with paired dissolved-solids readings where directed | A low TDS reading certifies every contaminant |
| Is a health-related target controlled? | Appropriate accredited laboratory analysis and professional interpretation | Clear water or a normal home strip is an all-clear |
| Is UV delivering its specified protection? | Exact system’s monitoring, water-quality and service requirements | A glowing lamp proves adequate dose |
Check the named model and claim using the certification guide before choosing the measurement. A filter may have several claims that need different evidence.
Plan the sample pair
Identify where each sample comes from and what treatment lies upstream. A hose bib may bypass the whole-house system; a refrigerator sample may have passed through an additional cartridge. Label the actual path, not just “before” and “after.”
Use the test manufacturer or laboratory’s collection, flushing, container and timing instructions. Do not apply a universal 30–60-second flush or rinse every bottle: some analyses require a specific stagnation protocol or supplied preservatives. If no appropriate untreated sample point exists, report that limitation rather than opening plumbing or bypassing treatment casually.
For a maintenance comparison, record the date, sample locations, test identity and expiry, units, equipment state, flow condition where relevant and time since service.
Check whether the result can resolve the claim
The method’s detection/reporting limit and precision matter. A coarse color block cannot reliably distinguish two concentrations that fall within the same band. “Not detected” means below the method’s limit, not zero. A test that cannot resolve the specified target cannot prove compliance with it.
If the incoming concentration is already below the method’s limit, an equally low outlet result does not demonstrate a removal percentage. If readings are comparable and suitable for the task, a calculated percentage is still specific to that sample pair and operating condition, not a certification of the product.
Respond to an unexpected result
Check sample identity, instructions, test condition and equipment status without dismissing a concerning result as user error. For a health-related concern, follow public-health advice and arrange appropriate confirmation. During a boil-water or other utility notice, follow the notice; do not self-clear the water with an ordinary home filter test.
For a maintenance issue, compare with the system’s defined service criteria and contact the supplier when the result is outside its limits. Use the maintenance checklist to record the action and required restart checks.
Method and limits
This is an original claim-to-measurement map informed by NSF’s treatment-standard distinctions . It supplies no universal sampling procedure, contaminant limit or service interval. No water was sampled for this article. For a private-well investigation, begin with the laboratory-panel guide .
Verify the exact certification claim before choosing a performance measurement .