Our method
How We Research Garment Steamers
No lab, no test bench, no units in a cupboard. Here is exactly what we do instead, and what that method cannot tell you.
By Scooter M., enthusiast, crease control · Published
We have not tested these steamers. We do not own them, nobody sent us one, and there is no lab behind this site. Everyone in this category says they tested twenty units. We have not tested any, and we would rather say so than pretend. What we did instead: read every manual and listing, put the published specifications side by side, work out what each machine actually costs to run, and publish the rubric so you can check the arithmetic yourself.
The five steps, in order
- 1Shortlist from the live categoryOnly currently-listed, currently-buyable models. Every product on this site was confirmed on a live Amazon listing during the research pass, and anything we could not verify is not here — including several brands with real search demand. The brand hub names them.
- 2Read the published specifications and record wherePower, tank capacity, stated heat-up, weight, voltage, dimensions and warranty, taken from each product's own listing, with the URL and the date. Every product page links back to the listing its figures came from.
- 3Drop the figures that are not figuresThree listings file a runtime in the heat-up field and one repeats the item weight as the tank capacity. Those are data errors, not specifications. We drop them and say which are missing, rather than filling in a plausible number.
- 4Score against the class, not against everythingA twelve-pound upright is not a failed handheld. Each figure is normalized against the range its own class actually spans, using the table below.
- 5Name what to skipEvery roundup on this site includes at least one product we would not buy, and every product page carries a 'who should not buy it' line. A page where everything is recommended recommends nothing.
The scoring rubric, in full
Four inputs, all published by the maker: power, tank capacity, stated heat-up and weight. Each is scored 0-10 against the range its class spans, then combined with the weights below. Nothing else feeds the score — not price, not brand, not review count, and certainly not an impression we did not form.
| Class | Power (W) | Tank (ml) | Heat-up (s) | Weight (lb) | Weights |
|---|---|---|---|---|---|
| Handheld | 700-1875 | 100-350 | 10-120 | 0.7-3.2 | power 0.3, tank 0.25, heat 0.2, weight 0.25 |
| Travel | 400-1100 | 60-200 | 10-60 | 0.5-2 | power 0.2, tank 0.2, heat 0.25, weight 0.35 |
| Standing | 1200-1800 | 1500-3000 | 30-60 | 8-13 | power 0.3, tank 0.35, heat 0.2, weight 0.15 |
| Commercial | 1200-1800 | 1800-3000 | 45-180 | 8-20 | power 0.3, tank 0.4, heat 0.3, weight 0 |
| Press | 700-1900 | 60-400 | 15-200 | 1-12 | power 0.35, tank 0.2, heat 0.2, weight 0.25 |
| Iron | 1200-1900 | 200-400 | 15-120 | 2-8 | power 0.4, tank 0.2, heat 0.2, weight 0.2 |
Power and tank score higher when larger; heat-up and weight score higher when smaller. A missing figure is dropped and the overall is the weighted mean of what the maker did publish.
The running-cost arithmetic
Energy in kilowatt-hours is watts divided by 1,000, times hours. Cost is that figure times your electricity price. We use the US average residential rate of 18.34 cents per kilowatt-hour (June 2026, US Energy Information Administration) and we cite it, because it moves every month. The full working is on the wattage page, so you can substitute your own rate.
Prices, and why they sometimes disappear
Every price on this site comes from a live data layer with a 48-hour freshness gate. If the snapshot is older than that, the number does not render and the button reads Check price on Amazon instead. That is deliberate: a wrong price is worse than no price, and a stale one is a wrong one. We never type a price into body copy.
Fabric claims
Every fabric claim on this site cites a fiber-property reference, a manufacturer's care guidance, or the FTC Care Labeling Rule — the federal rule that requires a garment maker to hold a reasonable basis for the instruction on the label. No fabric claim here rests on having steamed something, because we have not.
What this method cannot tell you
- Build quality. Whether a tank seal survives two years of thermal cycling needs a unit and time.
- Ergonomics. How a trigger feels on the fourth shirt is not in any specification sheet.
- Real-world steam quality. Two machines at the same wattage can produce noticeably different steam, and only measurement would show it.
- Failure rates. We can read a published warranty term; we cannot tell you how often it gets claimed.
Where that kind of judgment is what you need, the publishers who genuinely buy and test are worth your money. We link to them rather than paraphrasing their results, and we will not reproduce paywalled ratings we did not pay for.
Who writes this
One person: Scooter M., an enthusiast who is genuinely interested in this and reads the manuals. No credential, no panel, no lab coat. More on the about page, and the editorial policy sets out the rules this site holds itself to.
How this is funded
Affiliate commission from Amazon, disclosed on every page that carries a link and explained in full here. It does not decide the ranking — the ranking comes from the rubric above, which is why the rubric is published. The clearest evidence is our linen verdict, where the honest answer is an iron and there is nothing on that page for us to earn from.
Questions people actually ask
+Do you test the steamers you recommend?
No. We have not handled any of them, and we say so on every page. We read the published specifications off the live listing, drop the data errors, do the arithmetic, and publish the rubric so the ranking is reproducible.
+How do you decide the score?
Four published figures — power, tank capacity, stated heat-up and weight — each normalized against the range its class spans, then combined with the weights published in the table above. A missing figure is excluded rather than treated as zero.
+Why do some products show no price?
Because the live price layer has nothing fresh for them. A 48-hour freshness gate removes any figure older than that and the button falls back to Check price on Amazon. A stale price is a wrong price.
+Do the affiliate links change your rankings?
No. The ranking comes from the published rubric, which is why the rubric is public. Our linen verdict sends readers to an iron, which earns us nothing, and that is the point.