What does a shower filter remove?

What it does remove

• Free chlorine — the core function. Independently measured reductions across the category ranged from 42 to 100 percent, and this is the one thing NSF/ANSI 177 actually certifies.
• Sediment, rust flakes and particulates — every multi-stage cartridge has a mechanical stage doing this, and it is the stage that eventually clogs and costs you pressure.
• Hydrogen sulfide odor — KDF media acts on it, which matters mostly on private wells.
• Some dissolved metals — KDF's redox chemistry acts on some of them. No shower filter is certified for metal reduction, because the standard does not cover it.
• Some disinfection byproducts — but only from units with real carbon-block volume. Two out of the independent test set.

What it does not remove

• Hardness — calcium and magnesium. Independent testing found no filter in the category reduced it. Why.
• Chloramine, from a KDF-only filter. It needs vitamin C or catalytic carbon. The detail.
• Fluoride. Not meaningfully, and nothing is certified for it. Why not.
• Bacteria and viruses. No shower filter is a microbiological purifier and none claims to be.
• Nitrate, arsenic and most regulated drinking-water contaminants. They are not in scope for this product class at all.
• PFAS, with one exception: the Weddell Duo publishes NSF/ANSI 53 PFAS reduction test data. PFAS.

Why the second list is longer than the marketing suggests

Because there is only one standard for this product class and its scope is narrow. NSF/ANSI 177 requires a certified filter to reduce a 2.0 mg/L free available chlorine challenge by at least 50 percent across its rated life, hold structural integrity to 125 psi, use materials with no intentional lead, maintain at least 1 gpm at 80 psi, and publish specified literature. That is the entire testable surface.

The standard is explicitly not a health effects standard. So when a listing says a filter removes lead, fluoride, heavy metals, bacteria and 99 percent of contaminants, none of those claims is being tested by anything — not because the claims are necessarily false, but because no certification in this category evaluates them. There is nothing to check them against.

Contact time is the constraint behind all of this

One number explains most of the second list. Water is inside a shower head for 0.10 to 0.14 seconds at typical flow — a figure published by a filter manufacturer on its own technical page. Standard activated carbon needs 25 to 40 seconds of empty-bed contact to reduce chlorine by adsorption. Ion exchange, which is what removing hardness requires, is slower still.

This is why the media that work in a shower are the fast chemical ones — KDF redox, calcium sulfite reduction, ascorbic acid neutralization — and why the claims that fail are the ones requiring adsorption or exchange. It is not a quality problem that better engineering will solve. It is a residence time problem, and the residence time is set by how a shower works.

Reading a claim, in three questions

1. Is there a standard and a body named? "Certified to NSF/ANSI 177 by IAPMO R&T" is checkable. "Tested to NSF standards" and "surpassing NSF-177 certifications" are not the same thing.
2. Is the media named, or just counted? A media list is a commitment. A stage count is not.
3. Does the claim involve a mechanism that fits in a tenth of a second? Redox, yes. Chemical neutralization, yes. Adsorption and ion exchange, no.