
You cannot clean a mattress, you can only clean its surface. It cannot be rinsed, it cannot be wrung out and it cannot be dried, which removes every mechanism that makes ordinary washing work. What is left is genuinely useful and much narrower than most guides admit: vacuum thoroughly, spot treat individual marks with the least liquid that will do the job, deodorise dry with bicarbonate of soda, and dry hard. This page gives you that method, the evaporation budget that keeps it safe, and honest answers on the two things everyone actually wants, which are the yellow patches and the smell.

Washing anything works through three steps: you wet it, you agitate it, and then you remove the water along with whatever was dissolved in it. A shirt gets rinsed and spun. A cushion cover gets rinsed and spun. A mattress gets none of that. It is 200 to 300 mm of foam or springs, sealed inside a quilted cover, with no drainage, no way to compress the water out, and no dryer on earth that fits it.
Everything sensible about mattress care follows from that one fact. Any liquid you put in is a liquid you have committed to evaporating through a single surface, and evaporation through a single surface is slow. This is why almost every mattress disaster story ends the same way: the stain came out and the mattress smelled of damp for a month.
Manage that constraint and the rest is easy, because the good news is that most of what makes a mattress unpleasant lives in the top few millimetres. Skin, dust, mite debris and settled body oil all sit at or near the surface, and a vacuum reaches all of it.

Here is the rule that should replace every vague instruction to use a damp cloth: never apply more liquid than can evaporate from the surface within about two hours.
Two hours is not arbitrary. It is roughly the window in which a lightly dampened surface in a ventilated room dries before moisture has time to migrate downward into the core rather than upward into the air. Stay inside it and you are cleaning a surface. Go past it and you are irrigating a sponge you cannot subsequently dry.
In practice that means a spray bottle rather than a pour, a cloth wrung until it is barely cool to the touch, and one mark at a time rather than one whole surface. If a patch still looks visibly wet after ten minutes, you have already used too much, and the correct response is a fan and patience rather than more product.
It also settles the steam-cleaner question permanently. A steam cleaner exists to drive moisture into a surface under pressure. On a tiled floor that is exactly right. On a mattress it is the one thing the structure cannot survive, and memory foam is the worst case of all because it holds moisture for days.

Roughly three hours end to end, most of which is waiting, and most of the real benefit arriving in the first twenty minutes.
| Step | What it does | Time |
|---|---|---|
| Strip, wash bedding at 60 degrees | Kills dust mites, which cooler washes do not | Runs in parallel |
| Vacuum surface, sides and every seam | Removes the skin, dust and mite debris that is most of the problem | 15 to 20 min |
| Spot treat individual marks only | Targets stains without wetting the whole core | 10 min plus dwell |
| Sieve bicarbonate of soda over everything | Neutralises acidic odour compounds, entirely dry | Several hours |
| Vacuum the bicarbonate off completely | Leaves no residue to grit up under the sheet | 10 min |
| Force drying: fan, windows, sunlight | The only drying mechanism available to you | 1 to 2 hours |
| Fit a waterproof protector | Stops you ever needing to do this again | 2 min |
Sunlight is worth singling out because it is one of the few pieces of folk wisdom here that holds up. Ultraviolet light is antimicrobial and the warmth accelerates evaporation, which makes an afternoon in the sun a genuinely effective finishing step on the one household item you cannot tumble dry. Getting light and air to the bed at all is a layout question as much as a cleaning one, and the storage that surrounds it usually decides whether the room can be aired properly in the first place.

The sequencing logic is the same as it is for upholstery, and if the same accident also reached the sofa, the fabric-code half of it is worked out in how to clean a couch. What changes on a mattress is that every option is bounded by the two-hour rule.
| Stain | Use | Key detail | Never |
|---|---|---|---|
| Urine, fresh | Heavy blotting, then enzyme cleaner | Press with bodyweight through a dry towel before any product | Ammonia cleaners, which smell like urine to a pet |
| Urine, dried | Enzyme cleaner, full dwell time | Uric acid crystals reactivate in humidity, hence the returning smell | Detergent alone, which lifts the colour and leaves the crystals |
| Blood | Cold water, cold cloth, repeat passes | Heat cooks protein into the fibre and makes it permanent | Warm water at any stage |
| Sweat, yellowing | Nothing that works, realistically | See the section below. It is oxidation, not soil | Peroxide and bleach, which degrade the foam and the fire barrier |
| Vomit | Scrape solids, then enzyme cleaner | Bicarbonate afterwards for the residual acidity | Soaking, which pushes the problem into the core |
| Wine, coffee | Blot from the clean edge inward | Two weak passes with full drying beat one strong one | Rubbing, which widens the tideline every pass |

This is the disappointing part, and every guide that dodges it is wasting your afternoon.
Yellowing is not dirt. It is body oil and sweat that has oxidised inside the fibre over years, the same chemistry that yellows the collar of a white shirt and the pages of an old paperback. Oxidation is a change to the material itself rather than something sitting on top of it, so there is nothing to lift off. The only agents strong enough to reverse it are oxidising bleaches, and hydrogen peroxide and chlorine bleach both attack polyurethane foam and can damage the fire-barrier layer that sits under the cover. You would be trading a cosmetic mark for a structural and safety one.
The honest conclusion is that a yellowed mattress which still supports you properly and does not smell is cosmetically marked and functionally fine. Put a protector on it and stop looking at it. If the yellowing bothers you enough to spend money, spend it on the replacement rather than on the peroxide.

Most mattress warranties exclude a mattress that carries any stain or soiling, and manufacturers inspect for exactly that when a sagging claim arrives. It is worth reading your own paperwork rather than taking that on trust, but the pattern is common enough that you should assume it applies until you have checked.
That turns a single spill into a much more expensive event than it looks. A mattress with a ten-year warranty against sagging is carrying real value, and one cup of coffee can remove it, months before the sag that would have been covered ever appears.
Which produces the most useful piece of arithmetic in this article. A waterproof protector costs somewhere around a fortieth of a mid-range mattress. If you think there is even a few percent chance of a single spill, a pet, a child or an illness over the life of the mattress, the protector has already paid for itself in expected value, before you count the cleaning afternoons it saves you. It is the rare piece of household advice where the maths is not close. For context on what the mattress itself sits inside as a budget line, what it costs to furnish a house puts the bedroom against every other room.

Three signals, and only the first is objective.
It sags. Strip the bed, lay a straightedge or a taut string across the width, and measure the gap at the deepest point. Manufacturers commonly treat somewhere in the region of 25 to 40 mm of permanent indentation as a warranty defect, which is a sensible threshold to borrow whether or not you are still covered. Sagging is structural, and no cleaning method addresses structure.
It smells from standing height. The same test that decides the verdict on a sofa applies here and is more decisive, because a mattress has more volume and less access. If you can smell it standing up in a ventilated room, the contamination is through the core, and surface treatment is not going to reach it.
You sleep better elsewhere. Unscientific and still the most reliable signal most people have. A week in a hotel or at a relative’s house that consistently produces better mornings has told you something that no amount of bicarbonate is going to change.
If the answer is replacement, measure the route before you order rather than after. The mattress version of that problem is worked through in will a mattress fit through the door, and the sofa calculator uses the same geometry if a bed frame is going the same way.
A new mattress usually arrives with a new bed frame, and a new bed frame quietly resets the whole room: the headboard height changes the wall behind it, the frame depth changes the walkways, and the colour changes what the curtains are doing.
Photograph the bedroom as it stands and use the AI furniture changer to swap the bed in place, or describe the room in the generator and test an upholstered headboard against a low wooden frame before either has a delivery date. Every interior in this article was made that way.
The layout half of the decision, which is where the bed can physically go, is derived in the bedroom layout guide, and the textiles that end up framing it are covered in how to measure for curtains. The rest of the maintenance cluster, including the sofa this method is adapted from and the blinds above the bed, sits in care and cleaning.

Only its surface. A mattress cannot be rinsed, cannot be wrung out and cannot be dried in any controlled way, which rules out every mechanism that makes ordinary washing work. What you can genuinely do is remove what sits in the top few millimetres by vacuuming, spot treat individual marks with the least liquid that will work, deodorise dry with bicarbonate of soda, and dry hard afterwards. Anyone promising to make an eight-year-old mattress like new is describing a replacement.
Blot as much as physically possible first, pressing with a dry towel and your bodyweight rather than dabbing. Then an enzyme cleaner, applied to the area the urine actually reached and kept wet for the dwell time on the label, typically 10 to 15 minutes. Enzymes digest the uric acid crystals that ordinary detergent leaves behind, and those crystals are why a urine smell that seemed gone in winter comes back on the first humid day. Then dry with a fan and open windows. If the urine went in fresh and in volume, accept that some of it reached a depth you cannot treat.
Almost never, because they are not dirt. Yellowing is body oil and sweat that has oxidised in the fibre, chemically closer to how a white t-shirt yellows at the collar than to a spill. The agents strong enough to bleach it, hydrogen peroxide and chlorine bleach, degrade polyurethane foam and damage the fire-barrier layer, so the treatment costs more than the stain does. A yellowed mattress that is otherwise supportive and does not smell is cosmetically marked and functionally fine.
It deodorises rather than cleans, and that is worth doing. Bicarbonate of soda neutralises acidic odour compounds and adsorbs some moisture, and crucially it does this dry, which makes it the only whole-surface treatment a mattress can safely receive. Sieve a thin even layer, leave it several hours, then vacuum it off completely. What it will not do is lift a stain or reach anything below the first few millimetres.
Vacuum it every time you change the sheets if you can, and give it the full treatment, bicarbonate and all, about twice a year. Rotate it head to foot every three months, and flip it as well if it is a two-sided mattress, which most modern ones are not. The washing that matters most is the bedding: sheets weekly, and the protector monthly at the hottest temperature its label allows.
It is the single worst idea in mattress care. Steam drives moisture under pressure into a structure with no drainage and no way to dry, and memory foam in particular holds it for days. You solve a surface problem by creating a mildew problem inside the core, where nothing you own can reach it. The same logic that permits steam on a hard floor rules it out here.
On most mattresses, yes, and it is worth reading your own terms before you assume otherwise. Manufacturers commonly exclude mattresses with any stain or soiling, on hygiene grounds, and they inspect for exactly that when a sagging claim comes in. The practical consequence is uncomfortable: a single spill can cost you a warranty worth several hundred, which is what makes a waterproof protector the highest-return purchase in this entire article.
When it sags, when it smells from standing height, or when you consistently sleep worse on it than elsewhere. The sag test is the objective one: strip the bed, lay a straightedge across it, and measure the dip at the deepest point. Most warranties treat somewhere around 25 to 40 mm of permanent indentation as a defect, which is a reasonable threshold to borrow even out of warranty. Sagging is a structural failure, and no cleaning method addresses structure.