Verified contaminant removal
verified_removal
70
max · 70%
Weighted measured removal across contaminant categories. Filters need published measurements to earn points here.
Filtered contaminant categoriesRemoval percentages
Scoring methodology · water filters
How filter scores work
Water filters are scored against four evidence rules. The biggest one is measured contaminant removal, with smaller credits for indexed lab data, recognized certifications and disclosed filter technology.
463
filters scored
214
with lab data
229
with certifications
454
cover 10+ categories
Earned out of possible
The filter rubric is 4 rules worth 100 possible points. Each rule contributes up to its maximum, and the product page shows the exact breakdown behind the 0-100 score.
No proof, no credit
If a brand does not publish a measurement, certification or technology detail, that signal earns zero. The score rewards evidence, not vibes.
The rules
4 rules · 100 points possible
Lab data indexed
lab_indexed
15
max · 15%
Credit for a full third-party lab report or published lab dataset that can be checked from the product evidence.
Lab report statusSources
Certified by a recognized body
certification_body
10
max · 10%
Recognized certification bodies such as NSF or WQA add confidence that performance claims were independently reviewed.
Certifications
Filter technology disclosed
filter_technology_disclosed
5
max · 5%
A small transparency credit for explaining the filtration media or system design instead of hiding the mechanism.
Technologies
What removal covers
The removal rule looks across the contaminant categories Oasis returned for each filter. Serious public-health categories like PFAS, heavy metals, radiologicals and microbiologicals carry the most weight.
PFASHeavy metalsRadiological elementsMicrobiologicalsFluorideVOCsPesticidesPharmaceuticalsMicroplasticsNanoplasticsDisinfectantsTHMsHAAsHerbicidesOther published categories
How to read a filter page
The score breakdown shows the four rule scores. The contaminant removal table below it shows which categories have measured percentages and which ones are incomplete or unknown.
Certifications and technologies are confidence signals, but they do not replace actual removal data. A filter with strong disclosure should show both.