Does Reverse Osmosis Remove Microplastics and Nanoplastics?
RO research is promising, but a membrane result is not a guarantee for every household unit. Here is how to separate useful evidence from unsupported removal claims.
This is an evidence explainer, not a review of a tested household RO unit. We distinguish treatment-plant findings, experimental membranes and model-specific certification. No RO system is ranked or sold through this article.
Does reverse osmosis remove microplastics and nanoplastics?
RO can contribute to removing plastic particles, but “reverse osmosis” alone does not prove that a household system removes all microplastics or nanoplastics. Evidence from a treatment plant or an experimental membrane cannot be turned into a universal removal percentage for an under-sink unit.
The useful buying question is: what did this exact complete system reduce, over which particle sizes, and under what conditions? A microplastics claim does not automatically extend to smaller nanoplastics. A result for salt removal, water flow or membrane fouling is not a nanoplastic-removal result.
If you want product options rather than the evidence behind RO claims, start with our water-filter comparison. This article explains how to evaluate the stronger claim that a filter also addresses nanoplastics.
How RO fits into a household water system
RO uses pressure to push water through a semipermeable membrane. It produces treated water, called permeate, and a separate reject stream, called concentrate or brine. EPA describes this process and its water-use tradeoffs.
A complete system can include filters before and after the membrane. That matters when interpreting a claim: a test on one membrane is not automatically a test of the water leaving the final tap. NSF's overview of RO systems makes that multi-stage construction clear.
For this guide, “nanoplastics” means plastic particles below 1 micrometer, following the convention in the 2024 bottled-water imaging study. Definitions and analytical cutoffs vary. Read the numerical size range in a test report rather than relying on “micro” or “nano” in a headline.
What the research actually tested
| Primary source | What was studied | What it cannot establish |
|---|---|---|
| Cai et al., 2022 | A full-scale reclaimed-water plant using a membrane bioreactor followed by RO. Overall microplastic removal rose from 93.2% after the bioreactor to 98.0% after RO; some fibers remained. | The 98.0% figure is for the integrated treatment train, not the RO stage alone, a home appliance, or nanoplastics. |
| RO fouling study, 2024 | How model polystyrene nanoplastics interacted with scaling and fouling during RO desalination. | A universal removal percentage for retail drinking-water systems. |
| Farahbakhsh et al., 2024 | An experimental modified RO membrane. Its abstract reports 98.4% sodium-chloride rejection and improved resistance to nanoplastic fouling. | That 98.4% of nanoplastics were removed. The percentage refers to salt. |
| Qian et al., 2024 | Micro- and nanoplastic identification in bottled water; possible sources during production and packaging were discussed. | A controlled before-and-after test proving how much a home RO unit removes or releases. |
The plant study is evidence that treatment incorporating RO can reduce microplastics in that setting. It also shows why a membrane's theoretical separation ability should not be treated as a zero-particle guarantee for a complete installation. Household performance needs evidence tied to the household system.
NSF/ANSI 58, microplastics claims and nanoplastics
NSF/ANSI 58 addresses RO systems. A certification mark is useful, but it is not a promise to remove every contaminant. NSF explicitly advises checking the particular reduction claims attached to a model.
NSF also provides listings for microplastics reduction under NSF/ANSI 401. Find the exact system and cartridge, then read the claim and associated performance documentation. A logo or a reference to “NSF materials” is not the same evidence as certification of the complete system for the contaminant you care about.
For a nanoplastics claim, ask for the tested lower size limit. A certificate or report covering one particle range should not be extended to smaller particles without supporting data. We do not infer a nanoplastic-removal percentage from a standard number, a membrane marketing diagram, or a different model in the same brand's range.
Seven things to request before buying an RO system
Use this as a document checklist. It helps compare the quality of the evidence without needing to decide which marketing phrase sounds most scientific.
- Exact model and filter configuration. Match the system number, membrane, replacement cartridges and any final polishing or remineralization stage. A report for a different cartridge is not a substitute.
- The original report and laboratory. Ask for the report date, method and certification listing where applicable. A cropped percentage in a sales graphic leaves out the conditions.
- Particle sizes and materials. Look for the smallest and largest particles measured and the polymers tested. One type of spherical test particle does not represent every environmental fragment or fiber.
- Sampling location. Was water collected immediately after a membrane, or after the tank, final filter and dispensing tap? Prefer evidence for the configuration you will actually use.
- Both incoming and outgoing measurements. A reduction claim needs a meaningful comparison. Look for units, starting concentration, repeat samples and blank controls that check contamination from sampling and analysis.
- Detection limits. “Not detected” means below the method's detection capability in that sample. It does not prove that no smaller or otherwise unmeasured particles exist.
- Conditions and age. Ask about pressure, flow, source-water conditions, flushing and how much water had passed through the system. A new cartridge's result does not establish performance throughout its service life.
Our editorial decision rule: if the seller cannot supply evidence for nanoplastics, record that capability as unverified. You can still evaluate the unit for other documented treatment needs. There is no need to promote an unsupported nano claim to justify a purchase.
Can plastic filter components also contribute particles?
Particle removal and possible release from components are separate questions. In the bottled-water study, the authors noted that polyamide—one of the detected polymers—is widely used in RO membranes, and discussed equipment and other production stages as possible sources. That is a source hypothesis, not a controlled attribution to a particular household filter.
To judge a complete unit, useful evidence would compare incoming water with water from the final outlet under specified conditions. The mere presence of plastic in a housing cannot establish that a system makes water better or worse overall. Likewise, a membrane's removal result cannot rule out changes elsewhere in the water path.
Keep the decision practical: request whole-system evidence, follow the specified flushing and replacement procedure, and avoid buying a unit on either a “zero plastic particles” promise or an unsupported claim that every plastic-containing filter increases contamination.
Water use and maintenance belong in the comparison
EPA reports that typical point-of-use RO equipment can send five or more gallons to the reject stream for each gallon treated. WaterSense-labeled models must limit that to 2.3 gallons or less per treated gallon. WaterSense addresses efficiency and performance criteria; do not treat it as a universal nanoplastics claim. Check EPA's current criteria.
Before choosing a model, price the replacement filters and membrane, check the required water pressure and installation space, and establish who will maintain it. Use the manufacturer's schedule and instructions for your configuration. A filter you cannot maintain is a poor fit even if its initial documentation is strong.
RO is not automatically necessary for every household. Start with your actual treatment needs and compare them with verified reduction claims. A simpler filter may meet those needs with less water waste. For the container you fill afterward, our steel bottle materials guide explains lid and straw tradeoffs; a different bottle does not filter the water inside it.
Reverse osmosis and plastic-particle questions
Does reverse osmosis remove 100% of microplastics?
A universal 100% claim is not supported by the sources reviewed here. Treatment incorporating RO can reduce microplastics, but a household result must be tied to the complete model, tested particle range and operating conditions.
Can reverse osmosis remove nanoplastics?
RO is relevant to nanoplastic separation, but the technology name does not verify a retail unit's performance. Ask for a complete-system test that identifies the nanoparticle sizes, measurement method and reduction result.
Does NSF/ANSI 58 certification prove nanoplastic removal?
The standard number alone does not establish that claim. Check the exact model's certified reductions and request the tested particle-size range before drawing a conclusion about nanoplastics.
Is a low TDS reading proof that water has no microplastics?
No. Total dissolved solids and plastic-particle counts describe different measurements. A low TDS result is not a microplastic or nanoplastic analysis.
Does the bottled-water study prove home RO filters release nanoplastics?
No. It identified particles in bottled water and discussed possible sources during production and packaging. It did not run a controlled before-and-after test of household RO units.
Sources and editorial method
Reviewed October 29, 2026. We compared primary research abstracts, the full bottled-water paper, and EPA and NSF guidance. This is not an exhaustive systematic review, an independent filter test, or a prediction of health outcomes. We have not identified a universal household nanoplastic-removal percentage.
- Cai et al. (2022): integrated membrane treatment at a reclaimed-water plant.
- Impact of nanoplastics on RO scaling and fouling (2024).
- Farahbakhsh et al. (2024): modified experimental RO membranes.
- Qian et al. (2024): single-particle imaging of bottled-water nanoplastics.
- NSF: standards and the limits of certification claims.
- EPA: point-of-use RO and WaterSense.
If you already have a fridge dispenser, our refrigerator-filter guide shows how to check the existing cartridge and system before choosing additional equipment.