Cat Litter Testing Methodology

The short version

We buy a bag at full retail price, fill a box with it, live with it for the length of a run, and write down what we can weigh or count. Dust, clumping, odor and tracking come off instruments on a bench in a laboratory in Montreal. How the litter holds up with a real cat using it comes from ordinary homes, living rooms and garages. Cost per day comes from what the bag cost and how long it lasted. Those five things become five scores out of ten on the review page, and everything else you read there is opinion built on top of them.

How long the testing ran

The whole programme took about two years, several hours a day, five days a week. 2,847 hours of it. The bench covered 17 litters plus one additive, Purrify. Each product ran for 30, 60 or 90 days before its review, and the run length is printed on every review page. The reviews themselves went up between October 2025 and March 2026. Every bag was bought at full retail price, and no maker was offered a place on a list in return for anything.

You will not find a start date or a finish date for the testing anywhere on this site. Nobody wrote them down, so printing one would mean making it up. The months above are when reviews were published, which is a different thing.

What the readings are taken with

Four measurements come off instruments on a bench, not off a guess and not off a nose. Here is what each one is taken with. Dust took two instruments, because counting particles in the air and weighing them are two different jobs.

• Dust particulates: an optical particle counter and a weighed filterThe counter watches a pour as it happens and counts particles. The filter catches airborne dust and goes on a precision scale, and that filter weight is the milligrams per pour we publish.
• Clump tensile strength: a digital force gauge on a motorised test stand
• 72-hour odor absorption: an electrochemical ammonia meter
• Tracking radius: a measured floor grid and a precision scale

We name the type of instrument and stop there. The make and model of each one were never written down, so putting a brand here would mean inventing it, and you will find no accuracy or detection limit quoted anywhere on this site for the same reason.

How each reading is taken

Dust. A fresh box is poured, and it takes two instruments to read what a pour throws up. An optical particle counter draws air through a chamber and counts the particles, live, as the litter falls. A filter catches the airborne dust off a pour, and the filter goes on a precision scale. The milligrams per pour you see on every review page are the filter weight. The counter is what watches a pour happen; it counts particles and cannot weigh them, so no number on this site is credited to it. It is a pour test, not a room test, so it tells you what hits the air while you fill the box and not what is still hanging there an hour later. The pour height and the volume of litter poured were not recorded, so a reading here is comparable across our own bags but should not be lined up against another lab's figure. 16 of 17 litters have a dust reading. The rest say not measured.

Clumping. On the bench, clump tensile strength is measured with a digital force gauge on a motorised test stand: the stand pulls a clump apart and the gauge records the force it takes to break it. The figure we publish is a simpler one. Clumps are lifted and dropped from 3 feet, and we count how many survive the drop whole. That share is the clumping reading on every review page, and the tensile figures are not in the table below. Where the run recorded how many clumps were dropped, the count sits beside the percentage, and 50 clumps is the figure our recorded runs used. Where the count was not written down, the table says not measured rather than assuming the same 50. 12 of 17 litters have a clumping reading.

Odor. Ammonia in the box, in parts per million, read with an electrochemical ammonia meter as the run goes on. Ammonia is the compound behind litter-box smell, and it is worth metering rather than sniffing, because a human nose goes blind to it within minutes. Some readings could only prove the level was above or below a mark, and those are flagged rather than sharpened. The height above the litter at which each reading was taken was not recorded. On the bench the meter runs a 72-hour absorption check, and as with clumping and tracking those bench figures are not the ones in the table below. What we publish is the run-long timeline, and most runs were read weekly rather than at 24, 48 and 72 hours, which is why the hourly columns are mostly empty. 5 of 17 litters have a full ammonia run.

Tracking. Tracking radius is measured on a marked floor grid: granules carried out of the box are collected at set distances and weighed on a precision scale. The figure we publish is again the simpler one, a count of granules inside a 12-inch ring around the box. It tells you what lands near the box, not what a cat carries across a house, and the distance column in our table is empty for every product.

Cost. The price we paid for the tested bag, divided across the days it lasted in that run. It is one household's cost per day, not a national average, and it moves with litter depth, cat count and how often the box gets scooped. 17 of 17 litters have a costed daily figure.

What the ammonia numbers mean

A reading at or under 5 ppm sits below the published odor threshold for most people, so a box at that level reads as clean. At 13 ppm and above our testers could smell the box from across the room. 25 ppm is the level we treat as uncomfortable for the cat that has to put its face in it. Those thresholds are fixed for the whole site, so a number on one review means the same thing as the same number on another.

Where the testing happens

Both a laboratory and ordinary homes, and dropping either half would misdescribe the work. The four bench readings are taken in a laboratory in Montreal, on the instruments listed above. That is what makes those numbers worth anything. The living with it happens in living rooms and garages: 38 households and 71 cats across the runs, added up from the count recorded against each product. That is where odor in real use, tracking around the box and how a litter behaves when a real cat digs in it get scored, and a bench cannot tell you any of it. Runs last weeks rather than days, because a litter that behaves for three days and fails in week three is the exact failure a shopper needs warning about. Every run is published under the editorial team named on the editorial standards page.

How a score is built

Each review carries five scores out of ten: dust, clumping, odor, tracking and value. There is no secret weighted formula behind them. Where a reading exists, it anchors the score, and a litter cannot score well on dust while its weighing says otherwise. Where no reading exists, the score is editorial judgement and the review says which. Scores change when a new reading changes the picture, and a changed score changes the comparison pages and roundups with it, because all of them read the same catalog.

What is not recorded, and what is not tested

The type of instrument behind each reading is above. What was never written down is the make and model of each one, and the shop each tested bag was bought from. Those cells in the table say not recorded, and they will keep saying it until the paperwork exists. We would rather show the hole than name a brand we cannot prove we used.

We also do not test how a litter feels under a cat's paws, whether a cat still accepts it a year later, how a litter behaves in a covered or automatic box, or any health effect on a cat or a person. None of those get a number here.

Our standing rules

• Every published review uses the same five scoring categories
• A reading that was never taken shows as not measured, never as a zero
• A detail that was never written down shows as not recorded, never as a guess
• No number is ever borrowed from a similar bag to fill a gap
• Bags are bought at full retail price, and no manufacturer pays for a place on a list
• Recommendations change when a new reading changes the verdict

Every reading we hold

Scores are our judgement. The readings under them are not. Here is the whole table, the same one published on the test data page, where you can also see the run dates, the household counts and the full ammonia timelines behind each row.

The dust column has a report of its own: every litter with a dust reading, ranked, averaged by material, and published as a file you may quote with credit. Open the 2026 dust report.