About
The label and the lab, on the same line
The temperature in a sleeping bag’s name is a marketing decision. Its comfort rating is a measurement — and across the 11 bags here that publish both, the measurement lands a median 11 °F warmer than the name. A bag sold as a 20 keeps a cold sleeper comfortable nearer 31. TrailSleep prints those two numbers next to each other on all 61 entries and calls the difference what it is.
Widest gap in the catalogue
Sea to Summit Spark 15
- 15 °F
- printed in the name
- 28 °F
- ISO comfort, from the lab
- +13 °F
- honesty gap
Why the gap exists at all
ISO 23537 hands a tested bag three temperatures, not one: a comfort rating, where a cold sleeper is warm; a lower limit, where a warm sleeper is curled up and hanging on; and an extreme figure that is a survival number and no kind of plan. Makers almost always name the bag after the limit, which is legitimate and also flattering, and the buyer reads the name as the comfort figure. Nothing here is being caught out — the figures on this site are the makers’ own. They are just rarely displayed together.
The narrowest gap in the catalogue belongs to the Marmot Trestles Elite Eco 20 at 4 °F, so this varies by maker rather than being an industry-wide fudge. It only resolves where a lab figure exists at all: 16 of the 32 bags and quilts here carry an ISO or EN rating, and 21 publish no comfort figure. Those entries show the maker’s own number with a label saying so, and no gap, because there is nothing to subtract from.
How things are filed
Three families — bags, quilts and pads — split across 6 types, and they are never sorted into one another, because a temperature rating and an R-value are not the same kind of number and a list that interleaves them is lying about both. A quilt is filed as a quilt even where its maker markets it as a bag, since the thing that decides how it sleeps is the missing hood and the missing underside, not the word on the label. Pads are split by construction — air, self-inflating, closed-cell foam — because that is what decides weight, packed size and how much noise the thing makes at two in the morning.
Family also decides which fields a page is allowed to show. A pad page carries no temperature rating and a bag page carries no R-value, rather than either one borrowing the other’s units to look complete.
Copied from the maker, converted to one set of units
Every specification is read off the maker’s own published spec table — the per-size weight chart, the fill weight, the girth measurements, the denier of the shell — and converted into one set of units so a Colorado quilt workshop and a Swiss mattress maker can share a row. Where a maker leaves a field out and a named third party has measured it, that measurement is used and its source recorded. 77 source URLs sit behind the 61 entries, one set per item.
29 of the 29 pads carry an ASTM F3340 R-value, which is why R-values can honestly sit in one column here: since 2020 every serious maker has measured them the same way. Before that each brand used its own method and the figures were not comparable at all.
Six figures worked out here
Everything else on the site is copied. These are derived from published figures, labelled as estimates wherever they appear, and their formulas are written out on how we compare and on the spec-sheet page.
- Honesty gap
- Degrees Fahrenheit between a bag's ISO comfort rating and the temperature printed in its own name. Positive means the label is colder than the lab.
- Warmth per ounce
- Degrees below a 60 °F night bought per ounce carried, measured against the lab comfort figure rather than the marketing one.
- Loft budget
- Fill weight multiplied by fill power, in thousands of cubic inches — the loft a down item can physically produce, which is where its warmth comes from.
- R per ounce
- ASTM F3340 R-value divided by pad weight: the pad's side of the same weight argument bags are judged on.
- Season band
- The season an R-value actually buys — summer, three-season, cold three-season or winter — on the boundaries the industry settled after ASTM F3340 arrived in 2020.
- Recommended pad R-value
- The minimum R-value the bag-to-pad table pairs with a given comfort rating, because insulation crushed under your body stops working and the pad carries the whole underside.
What owners said, and where it is kept
A spec table gives you the denier of a pad’s face fabric. It does not tell you the pad crackles every time you roll over, or that a zipper snags on the draft tube once it is cold, or that a quilt’s straps let go on a narrow pad. That comes from public owner discussion and long-term write-ups, and it is collected for 44 of the 61 entries.
It lives under its own heading on the page, never blended into the specification, so it is always obvious which figures are the maker’s and which sentences are somebody’s night out in the cold. The remaining 17 entries say in as many words that no owner discussion has been read for them. Nothing here claims owners said something about a product whose owners nobody has read.
The figures no maker publishes
A blank is information. No published comfort rating means no honesty gap and no warmth-per-ounce figure, and the page says so rather than filling the space with the marketing number wearing a lab coat. No published fill weight means no loft budget. No published shoulder girth means no girth row, not a guess from a photograph.
Star ratings work the same way: 0 of 61 entries carry one, because a rating appears only where a named non-Amazon retailer or a maker’s own store publishes an average that can be pointed at. Amazon ratings are never republished here and no score is ever estimated, which is also why this site emits no aggregate rating in its structured data. Prices are approximate snapshots from the last catalogue update and are labelled that way everywhere they appear.
Two people, and no gear closet
Two people, working on this alongside other sites. There is no office, no test lab and nobody here who has slept in all 61 of these — and the site never claims otherwise, which is the reason there are no picks and no awards on it. Ordering is arithmetic: coldest lab comfort rating first for bags, highest R-value first for pads, with the rule printed at the top of every ranked list and applied to the whole catalogue with nothing held back.
No money changes hands with any maker. Nothing here is paid placement and no brand has ever been shown a page before it shipped. TrailSleep carries no retailer links at all — no buy buttons, no affiliate tags, no commission on anything you do after leaving — so no ordering on this site can earn anyone anything. Manufacturer URLs exist in the data as the source a figure was read from, and are not rendered as links. What that costs and what it collects is on the privacy page.
When the bag and pad data was last checked
New models, discontinued ones, revised R-values and price moves are folded in as they happen rather than on a schedule, because sleep gear does not release on one. The catalogue was last updated September 2026, and that date — the data file’s own — is what the footer shows and what every sitemap entry carries, rather than the day you happened to load the page. If a figure here has gone stale or was wrong to begin with, tell us and it gets fixed.
Start here
Every entry shows its own sources, its own blanks and, where there is a lab figure, its own gap.