Mattress Firmness & ILD Ratings

ILD — Indentation Load Deflection — is the mattress industry’s primary firmness specification. It appears on product pages, in comparison articles, and in review rubrics as if it were a complete description of how a mattress will feel and perform. It is not. ILD is a single-point measurement taken under specific laboratory conditions that captures one dimension of foam mechanical behaviour — the elastic response at room temperature at a fixed strain rate. Everything that determines actual sleep performance — viscoelastic conformance over time, temperature sensitivity, layer interaction, long-term durability — is invisible to ILD alone. This article explains what ILD measures, where it falls short, and how to use it alongside the specifications that fill the gaps.

1. What ILD Actually Measures

The test method

ILD is measured according to ASTM D3574 Test B. A foam sample — standardised at 100 mm thickness — is placed on a rigid platen. A circular indenter plate (323 cm² surface area, approximately 20 cm diameter) is pressed into the foam surface at a controlled rate. The force required to achieve 25% indentation of the original sample thickness is recorded. This force, expressed in Newtons (N) or sometimes pounds-force (lbf), is the ILD value.

Some specifications also report ILD at 40% indentation (a deeper compression that engages the foam’s non-linear stiffening behaviour) and the “support factor” — the ratio of ILD at 65% to ILD at 25% — which indicates how quickly the foam stiffens under increasing load. A high support factor means the foam provides more resistance as compression increases, which is generally desirable for supporting heavier body regions without bottoming out.

What the number tells you

ILD tells you the force required to compress the foam to 25% of its thickness with a standardised indenter. Higher ILD means more force is required — the foam is firmer. The approximate firmness language used across the industry maps to ILD ranges roughly as follows:

• Below 10N: extra soft — rarely used in mattress comfort layers; more common in specialty medical applications.
• 10–20N: soft — used in plush comfort layers for side sleepers, particularly lighter individuals.
• 20–30N: medium to medium-soft — the most common range for memory foam comfort layers in the consumer market.
• 30–45N: medium-firm — suited to back sleepers and heavier side sleepers; common in latex comfort layers.
• 45–60N: firm — typically used in support transition layers or as comfort layers for stomach sleepers and heavy individuals.
• Above 60N: extra firm — generally used in support cores rather than comfort layers.

These ranges are approximate and not standardised across manufacturers. A “medium” mattress from one brand may have an ILD that falls in another brand’s “medium-firm” range. The language is marketing; the number is the specification.

2. What ILD Does Not Tell You

The limitations of ILD as a performance specification are extensive and practically important. Understanding them prevents over-relying on a single number to predict sleep experience.

ILD does not characterise viscoelastic behaviour

The ASTM D3574 test is conducted at a specific, relatively fast compression rate. It captures the elastic component of the foam’s response — the immediate, rate-independent stiffness. It does not characterise the time-dependent viscous component that drives stress relaxation and progressive conformance in viscoelastic foam.

Two foams with identical ILD values can have dramatically different loss factors (tan δ) — one may be a true slow-recovery memory foam with high energy dissipation, the other a faster-recovering hybrid with low energy dissipation. Their pressure distribution performance over an 8-hour sleep period will differ substantially, but their ILD ratings will be identical. As covered in the Viscoelastic Mechanics article, the loss factor and resilience are the specifications that characterise this behaviour — not ILD.

ILD is measured at a single temperature

The standard test temperature is 23°C. Viscoelastic foam’s mechanical properties are strongly temperature-dependent — the same foam measured at 16°C will show a significantly higher ILD than at 23°C, and at 35°C (approximate skin surface temperature after sustained contact) it will show a lower ILD. The ILD on the product specification sheet tells you how the foam performs at 23°C in a laboratory; it tells you less about how it will perform in your bedroom in January or July.

This limitation is more significant for memory foam than for HR foam or latex. As discussed in the Thermoregulation article, memory foam’s temperature sensitivity is a design feature — but it means the ILD rating understates how firm the mattress will feel in cold conditions and overstates it in warm conditions.

ILD is a surface measurement, not a system measurement

The ASTM test measures a single foam sample in isolation. A mattress is a stratified system of multiple layers with different mechanical properties. The ILD of the top comfort layer does not predict the feel of the assembled system, because the layers interact mechanically: the stiffness of the comfort layer is modified by the constraint provided by the transition layer and support core beneath it.

A soft comfort layer over a very firm support core will feel firmer than the same comfort layer over a medium support core, because the comfort layer bottoms out — reaches the point where the much stiffer core dominates the response — sooner. The effective firmness experienced by a sleeping body is a property of the system, not of any individual layer, and it cannot be predicted from the comfort layer’s ILD alone.

ILD does not predict durability

ILD is a measure of initial stiffness. It tells you nothing about how that stiffness will change over time. As covered in the Foam Degradation article, long-term performance is predicted by density — a specification that ILD is frequently substituted for in marketing materials, to the detriment of informed purchasing decisions. A high-ILD, low-density foam will feel firm initially and degrade rapidly. A lower-ILD, high-density foam will feel softer but maintain its properties for years longer.

3. The Firmness Scale Problem

The mattress industry uses a 1–10 firmness scale — 1 being extremely soft, 10 being extremely firm — in product descriptions, comparison sites, and reviews. This scale has no standardised definition. It is not anchored to ILD values. It is not calibrated across manufacturers. It reflects a combination of subjective tester impressions and marketing positioning, and it varies substantially between sources.

A mattress rated 6/10 by one reviewer may be rated 7/10 by another reviewer using a different reference point. A brand that rates its medium mattress as 5/10 may be using a different calibration than a brand that rates its medium as 6/10. The same mattress can receive different firmness ratings from reviewers of different body weights, because heavier individuals compress the comfort layer further and engage a stiffer response region.

The 1–10 scale is a useful communication shortcut for subjective feel comparisons within a single reviewer’s consistent framework. It is unreliable for cross-brand comparisons and meaningless without knowing the reviewer’s body weight, sleep position, and reference mattress. ILD, despite its limitations, is more useful for cross-brand comparison because it is at least anchored to a standardised measurement — even if that measurement is incomplete.

4. Firmness by Sleep Position and Body Weight

With the limitations of ILD established, it remains useful as one input into the selection framework — combined with sleep position and body weight to narrow the appropriate range.

Side sleepers

Side sleeping generates high peak pressures at the shoulder and hip. The comfort layer must be soft enough to allow these bony prominences to sink into the surface, increasing contact area and reducing peak pressure. For most side sleepers, comfort layer ILD in the range of 15–25N provides the necessary conformance. Heavier side sleepers (above 90 kg) may need somewhat higher ILD (20–30N) to avoid bottoming out the comfort layer — sinking through it to the firmer support core — which would negate the pressure relief benefit.

Back sleepers

Back sleeping distributes load more evenly across the dorsal surface and requires lumbar support to fill the lordotic gap. The comfort layer can be firmer than for side sleeping — typically 25–40N — without generating problematic peak pressures, because the bony prominences in the supine position (sacrum, thoracic spine) are less pronounced than the shoulder and hip in lateral position. The support core firmness matters more for back sleepers than for side sleepers.

Stomach sleepers

Stomach sleeping requires the firmest surfaces — typically 35–50N comfort layer ILD — to prevent the hips from sinking and the lumbar spine from extending excessively. The pressure concerns in prone sleeping are at the sternum and anterior iliac crests; these are managed by surface firmness rather than conformance. Note that no mattress specification can address the cervical rotation inherent in prone sleeping.

Combination sleepers

Sleepers who move between side and back positions during the night need a surface that performs adequately in both — which typically means medium firmness (25–35N) with good pressure distribution performance in the side-lying position. Hybrid designs with zoned support can partially optimise for both positions simultaneously.

Body weight scaling

Body weight scales the effective firmness experienced on any given surface. The same foam under a 60 kg sleeper produces different compression depth and stiffness engagement than under a 100 kg sleeper. As a general guideline:

• Below 60 kg: prefer ILD 1–2 grades softer than the standard recommendation for your sleep position.
• 60–90 kg: standard ILD recommendations apply.
• Above 90 kg: prefer ILD 1–2 grades firmer than standard, and prioritise high-density foam to manage accelerated compression set.

5. Building a Complete Specification Framework

ILD is one specification among several. A complete evaluation framework for foam mattress selection uses ILD in conjunction with the following:

Foam density (kg/m³)

The primary durability predictor. For viscoelastic comfort layers: minimum 40 kg/m³ for acceptable durability, 50+ kg/m³ for good durability. For support cores: minimum 28 kg/m³ for HR foam. This specification should be obtained for each layer separately — not as a system average.

Resilience or loss factor

Characterises the viscoelastic vs elastic character of the foam. Low resilience (below 15%) indicates genuine memory foam behaviour with high energy dissipation and slow recovery. High resilience (above 50%) indicates HR foam or latex character with fast recovery. This specification determines motion isolation, position-change responsiveness, and temperature sensitivity.

Support factor (ILD65/ILD25)

The ratio of 65% ILD to 25% ILD. A support factor above 2.5 indicates a foam that stiffens progressively under increasing load — desirable for preventing bottoming out under heavy loads and for providing differential support across body regions. Memory foams typically have lower support factors than HR foams or latex.

Layer structure and thickness

The comfort layer thickness determines how much the body can sink before reaching the transition layer and core. A 5 cm comfort layer and a 10 cm comfort layer with identical ILD will feel different to a heavy sleeper because the heavier sleeper may compress through the thinner layer to engage the stiffer substrate. Comfort layer thickness should be matched to body weight: heavier sleepers generally benefit from thicker comfort layers (8–12 cm) to avoid bottoming out.

6. Testing for Yourself: Beyond the Specification Sheet

Specifications narrow the field; direct testing confirms the selection. A few principles for making in-store or at-home trial testing more informative than a thirty-second impression:

• Test in your primary sleep position, not sitting on the edge. The pressure distribution in side-lying is completely different from seated, and the thirty-second sit-test tells you nothing relevant about sleep performance.
• Allow time for stress relaxation. On a memory foam surface, the relevant feel is not the initial elastic response but the state after 5–10 minutes of sustained contact. If testing in a store, lie still for at least 5 minutes before assessing conformance.
• Test lumbar support in the supine position. Slide a hand under your lower back while lying supine. There should be just enough space for your hand to fit — indicating the mattress is filling the lumbar gap without pushing the spine into flexion.
• Assess shoulder sinkage in side-lying. Your spine should be approximately parallel to the mattress surface — visible to a partner or in a mirror. If your shoulder is significantly higher than your hip, the mattress is too firm. If your hip is significantly higher, it is too soft for your body geometry.

Summary

ILD is a useful but incomplete specification. It measures the elastic response of an isolated foam sample at a standardised temperature and compression rate — capturing one dimension of a multi-dimensional performance profile. It does not characterise viscoelastic behaviour, temperature sensitivity, layer interaction, or long-term durability. Used in isolation, ILD cannot predict sleep performance; used alongside density, resilience, support factor, and layer structure, it becomes a meaningful input to an informed selection framework.

The firmness-scale language prevalent in mattress marketing adds a layer of subjective variability that further reduces its utility for cross-brand comparison. When a specification sheet is available, the ILD number is more useful than the firmness label. When neither is available, the absence of specification data is itself informative about the manufacturer’s transparency.

Next in this series: Spring Systems — Coil vs Pocketed — the mechanical differences between Bonnell, offset, and individually pocketed coil systems, and how the spring architecture interacts with the foam comfort layers above it.

The Sleep Mechanic is a materials engineer with hands-on R&D experience in cushioning materials and viscoelastic polymers. Sleep Science Lab applies materials engineering analysis to sleep surfaces — because “it feels comfortable” is not an explanation.