Mattress Firmness Numbers, Explained: What the 1–10 Scale Actually Measures

The 1-to-10 firmness scale measures no standard at all — it is each brand's own description of its own bed, with no shared test, load or panel behind it, which is why one maker's 'firm' and another's 'medium' can be the same 6.5 out of 10.

The 1-to-10 scale measures a feeling, assigned by the seller. There is no industry standard behind it: no shared machine, no agreed load, no common panel of sleepers and no regulator that checks whether one brand's 6 is another brand's 6. Each maker places its own bed on its own scale, usually by internal testing and judgment, and prints the result next to a word — soft, medium, medium-firm, firm. The number is therefore useful only inside one brand's own line, where it at least ranks that maker's beds against each other. Across brands it cannot be compared directly, and our mattresses ranking shows why in its own table: beds described by their makers with different words land on the same number, and beds carrying the same word land a full point apart. What the scale actually measures is marketing position. What a buyer needs — how the bed will hold a particular body, in a particular position — has to be read from other evidence.

A scale with no scale-maker

Most product numbers a shopper meets have an owner: an agency, a standards body, a test protocol that another laboratory could repeat. The mattress firmness number has none. Nothing in US law defines "medium-firm," no federal test assigns the 1-to-10 figure, and brands are not required to disclose how theirs was produced — some use pressure mapping or internal panels, some use the judgment of their own product teams, and the method rarely travels with the number. The consequence is visible without leaving our own ranking. One all-foam bed is listed as "Firm per maker specs" while the maker's own FAQ puts it at 6.5 out of 10. Another hybrid is marketed with a firm-leaning feel and is read by testers at about 6.5 out of 10. A third maker's "Medium" latex hybrid tests firmer than either, at about 7 out of 10. Read literally, Firm equals 6.5, medium-firm equals 6.5, and Medium equals 7 — the words and the numbers are each brand's private language, and a shopper who lines up three brands' "firm" beds is lining up three unrelated claims.

The one firmness number that is a real measurement

A standardized firmness measurement does exist — one level down, inside the foam. Under ASTM D3574, Test B1, indentation force deflection is the force a flat circular indenter of 50 square inches must apply to compress a flexible foam sample by 25% of its thickness, read after the deflection is held. The higher the force, in pounds, the firmer that foam layer. It is a repeatable laboratory figure, and it exposes the shopping scale's limits by contrast: the test rates a single foam layer, not a finished mattress, and a finished bed stacks several layers over coils or a support core, so no layer's number sums into the whole bed's feel. Almost no mattress listing publishes it. The buyer is offered the 1-to-10 description, which cannot be re-tested, in place of the layer measurement, which could be — and the substitution is worth noticing whenever a number is presented as though it were a specification.

What the clinical evidence actually used

The strongest evidence in this category is usually quoted as "medium-firm wins," and its methods show how far the trial is from the store shelf. In the randomized, double-blind trial published in The Lancet in 2003, Kovacs and colleagues assigned 313 adults with chronic non-specific low-back pain to firm or medium-firm mattresses for 90 days. At 90 days the medium-firm group had better outcomes for pain in bed (odds ratio 2.36) and for disability (odds ratio 2.10) than the firm group, and the authors concluded that a mattress of medium firmness improves pain and disability in that population. But the trial's firmness was rated on a scale developed by the European Committee for Standardisation on which 1.0 is the firmest and 10.0 is the softest — the reverse direction of the US shopping scale — and its two beds sat at 2.3 (firm) and 5.6 (medium-firm) on that scale. Those numbers cannot be carried onto a retailer's 1-to-10 chart, where higher means firmer and the middle of the chart means something else. The trial supports a direction — medium-firm rather than firm, for that group of back-pain patients — not a shopping number, and any brand that cites it as proof of its own "6.5" is borrowing authority the study never issued.

Firmness is not support, and neither is the sleeper constant

Two beds with the same honest feel number can still treat a body differently, because firmness describes the surface's first resistance while support describes whether the spine stays in line once the body has settled — and the second depends on the sleeper. The same bed compresses less under a lighter sleeper and more under a heavier one, so the heavier sleeper meets the firmer layers deeper in the bed and reads the mattress as softer at the shoulders and harder at the hips than the lighter sleeper does. Position moves the target as well: our ranking's own guidance is that side sleepers — especially those under 150 pounds — should lean one step softer than the label, while stomach sleepers and heavier sleepers should lean one step firmer, and that for most back and combination sleepers medium-firm is the safe buy. A number that changes meaning with the body on it is a starting vocabulary, not a verdict; the trial period, which in our ranking runs from 90 nights to a full year, is the only test conducted on the actual sleeper, which is why its length and return terms deserve more weight in the decision than the firmness figure.

How to use the number without being used by it

Used inside one brand, the scale does its one honest job: it orders that maker's own beds from softer to firmer, so a sleeper who knows a maker's medium was too soft can step to that same maker's firmer build with some confidence. Used across brands, it should be demoted to a rough neighborhood — most beds sold today cluster in the middle band the trade calls medium to medium-firm — and the decision moved to evidence that travels: the bed's construction and layer order, the trial length and the cost of a return, the warranty's sag threshold, and the maker's own answers when the label and the FAQ disagree. When a maker's specification sheet and its FAQ give different accounts of the same bed's firmness, believe that the disagreement itself is the information: the number was never measured tightly enough to argue with. Our methodology applies that rule across the Sleep rankings — a claim counts as written, for the exact model, on the maker's own pages — and firmness is the category where the rule earns its keep most often.

Sources

• Kovacs FM, Abraira V, Peña A, et al. — Effect of firmness of mattress on chronic non-specific low-back pain: randomised, double-blind, controlled, multicentre trial, The Lancet 2003; 362: 1599–604 (PubMed 14630439): 313 adults with chronic non-specific low-back pain and no referred pain; firmness rated on the European Committee for Standardisation H(s) scale from 1.0 (firmest) to 10.0 (softest); firm mattresses at H(s) 2.3 and medium-firm at H(s) 5.6; at 90 days the medium-firm group had better outcomes for pain in bed (odds ratio 2.36 [95% CI 1.13–4.93]) and disability (2.10 [1.24–3.56]); a mattress of medium firmness improves pain and disability among patients with chronic non-specific low-back pain: https://pubmed.ncbi.nlm.nih.gov/14630439/
• ASTM D3574, Test B1 — indentation force deflection, as described in the IFD/ILD testing guide: indentation force deflection is the force needed for a flat circular indenter foot of 50 square inches (323 cm²) to compress flexible foam by 25% of its thickness, read 60 seconds after the deflection is reached, with a 65% reading also taken; the higher the figure, the firmer the foam; the number is meaningful only together with the deflection percentage, thickness and specimen, and it rates the foam specimen, not a finished product: https://biopdi.com/indentation-load-force-deflection-test/
• AmRank — mattresses ranking: firmness descriptions are each maker's own — one all-foam bed is "Firm per maker specs" while its maker's FAQ publishes 6.5/10; a second hybrid (DreamCloud) markets a firm-leaning feel that testers read as about 6.5/10, and a luxury-firm hybrid (WinkBed) sits at about 6.5–7/10; a "Medium" latex hybrid reads medium-firm at about 7/10; a "Medium" hybrid is balanced at about 6/10; ranking guidance holds that most back and combination sleepers are safest with medium-firm, side sleepers (especially under 150 lb) should lean one step softer than the label and stomach and heavier sleepers one step firmer; trials in the ranking run from 90 nights to 365 nights: https://americanrankinginstitute.org/rankings/sleep/mattresses/