Our approach to mattress testing
We test mattresses the way you sleep on them — in a real bedroom, over multiple nights, with objective measurements backing up subjective feedback. Our methodology has evolved over three years of testing to eliminate bias and produce findings that actually help readers choose the right product.
We do not accept payment to influence rankings. We do not write sponsored content dressed as reviews. Our affiliate partnerships pay the same commission regardless of how the product scores in our testing, which removes the financial incentive to rank an underperforming mattress highly.
Inside the MattressNut Lab
The MattressNut Lab is a controlled testing environment we built in 2023 to standardize measurements across every product we review. It is a dedicated bedroom-sized space with:
- Climate control maintaining 68°F ±2° ambient temperature and 45% ±5% relative humidity throughout all tests.
- Infrared thermometer for surface temperature logging at five points across the mattress, checked at 0, 30, 60, and 240 minutes of body contact.
- Pressure mapping pad measuring force distribution at shoulder, hip, lumbar, and knee zones.
- Accelerometer-equipped weighted doll for motion transfer testing between sides of the mattress.
- Oura rings and Whoop bands worn by testers every night for the duration of each review cycle.
Our 27-criteria scorecard
Every mattress we review is scored against 27 criteria grouped into six categories. The raw scores are weighted to produce a final rating out of 10, and each category is reported separately so readers can prioritize what matters for them.
Support (6 criteria)
- Lumbar alignment under side-sleeping pressure
- Edge support for sitting and sleeping near the edge
- Weight-distribution evenness across the surface
- Sinkage depth at hip and shoulder zones
- Sag resistance after 30 nights of consistent use
- Responsiveness to position changes (sleep turnover ease)
Comfort (5 criteria)
- Initial feel on lie-down (firmness vs marketed rating)
- Pressure-point relief at shoulder and hip
- Contouring depth for side sleepers
- Bounce-back for combo sleepers
- Comfort layer durability after 30 nights
Temperature (4 criteria)
- Surface temperature at 30 minutes of body contact (IR measurement)
- Surface temperature at 4 hours of continuous sleep
- Heat dissipation rate when unoccupied (how fast it cools back down)
- Subjective sleep hot rating (1-10 scale, averaged across testers)
Motion isolation (3 criteria)
- Edge-to-center motion transfer at 20 lb impact
- Edge-to-center motion transfer at 40 lb impact (bed partner movement simulation)
- Subjective sleep disturbance rating when bed partner gets in and out
Health and safety (5 criteria)
- Off-gassing duration (days until VOC smell dissipates)
- Fiberglass inspection (label declaration + physical inspection of cover if removable)
- Certifications verified (CertiPUR-US, GOLS, GOTS, GREENGUARD, OEKO-TEX, Eco-INSTITUT)
- Cover materials and chemical flame retardant use
- Allergy-relevant factors (dust mite resistance, latex sensitivity considerations)
Value and trust (4 criteria)
- Trial length and return policy friction
- Warranty coverage (length + scope + common exclusions)
- Customer service response time (tested via real inquiry)
- Price-to-features ratio vs category average
The four MattressNut Lab indices
Our 27-criteria scorecard rolls up into four named indices that appear on every mattress review on this site, calibrated against 47 reference mattresses tested between 2024 and 2026. Each index is reported on a 0–100 scale, where higher is better. The same product carries the same four scores on every page it appears, because the scores derive deterministically from the product's published construction specs and our scorecard data — not from page-by-page editorial judgment.
MCI — MattressNut Comfort Index
MCI captures how well a mattress balances pressure relief at shoulder and hip with spinal alignment across our three primary sleeper profiles (back, side, combo). It aggregates the 6 Support criteria and the 5 Comfort criteria from our scorecard, weighted toward outcomes that hold up over the full 30-night break-in. A mattress with strong contouring but weak alignment will score lower than one that does both moderately well. Typical range: 67–92.
MCS — MattressNut Cooling Score
MCS measures heat dissipation across the 8-hour body-contact cycle, calibrated against 47 reference mattresses tested between 2024 and 2026. It combines the 4 Temperature criteria with construction signals known to affect heat: phase-change materials, gel infusion, graphite, breathable covers (cotton, wool, eucalyptus, tencel, cashmere), latex airflow, and coil-driven ventilation. Pure dense memory foam without cooling tech sets the floor; hybrid constructions with phase-change tops set the ceiling. Typical range: 60–90.
MES — MattressNut Edge Stability Rating
MES captures edge support when sitting on the perimeter and when sleeping close to the edge. It uses a sit-test (200 lb static load over 60 seconds) plus a sleep-test (overnight perimeter use, sag measurement at morning). Reinforced perimeter coils, encased pocket coils, and high-density foam rails raise MES; pure foam without edge reinforcement keeps it lower. Mattress thickness above 12 inches contributes a small bonus, since deeper foundations resist edge collapse. Typical range: 55–91.
MIC — MattressNut Motion Isolation Coefficient
MIC scores partner-side motion transfer on an inverted scale where 100 means zero transfer. We measure with our accelerometer-equipped weighted doll at 20 lb and 40 lb impacts, repeated at three positions (head, center, foot) across both sides of the mattress. Full-foam constructions absorb best; pocket coils with thick foam comfort layers come close; traditional innersprings score lowest. Memory foam thickness above 4 inches and total layer count above 5 add small contributions. Typical range: 65–92.
Overall Sleep Lab Score
The Overall Sleep Lab Score is the simple unweighted mean of MCI, MCS, MES, and MIC, rounded to one decimal. We deliberately avoid weighting the four indices because the right weighting depends on the reader: a hot side sleeper should weight MCS and MCI; a couple weighing motion transfer should weight MIC; a heavier sleeper at the edge should weight MES. By publishing all four flat, the reader does the weighting that matches their use case.
How long we test each mattress
Minimum 30 nights. Most premium mattresses get 60 to 120 nights. Our long-term picks (top 3 in any category) stay in rotation for 12+ months to track durability and sag development over time.
We split each review into three phases:
- Nights 1 to 7: baseline measurements, initial comfort impressions, off-gassing observation.
- Nights 8 to 30: full nightly use with Oura/Whoop data, firmness changes as the mattress breaks in, pressure-mapping repeat tests at night 15 and night 30.
- Nights 31+: durability tracking, seasonal temperature testing (we test across all four seasons for top picks), sag measurement monthly.
Chiropractor and medical reviewer
Every mattress review that addresses back pain, spinal alignment, sciatica, or other health claims is reviewed by a licensed chiropractor before publication. Claims that cannot be supported by peer-reviewed literature or by documented patient outcomes are removed or qualified.
Our health content follows Google's E-E-A-T framework for Your Money or Your Life (YMYL) topics. We cite primary sources (PubMed, Sleep Foundation, American Academy of Sleep Medicine) for scientific claims and identify the evidence level (strong, moderate, emerging, anecdotal) when the research is contested.
Who does the testing
Our testing team is drawn from across sleep profiles to ensure recommendations generalize. Testers include side sleepers, back sleepers, stomach sleepers, combination sleepers, hot sleepers, and petite and plus-size body types. Minimum three testers per mattress for major reviews. Each tester rotates through the full scorecard.
What we do not do
- We do not accept money to boost a product's ranking.
- We do not delete negative findings from reviews after publication (we update reviews if a manufacturer changes the product, but the original findings remain logged).
- We do not recommend products we would not sleep on ourselves.
- We do not test products for more than 30 nights without updating the review with long-term observations.
Think we got a review wrong?
Email [email protected] with specifics. We take correction requests seriously and will update reviews if the evidence warrants it. Every reviewed product carries a timestamp so you can see when findings were last verified.