Compare bags on the comfort figure, and treat the lowest number on the label as a laboratory output rather than a plan. All three come from the same test, and they are not three grades of the same thing.
Where the three figures come from
Bags tested to EN 13537 or its successor ISO 23537 are measured on a heated mannequin under controlled conditions, and the standard calculates more than one value from that measurement. Makers who test to the standard publish what it yields.
That is why the figures appear together, and why they are only meaningful when the standard is named beside them. A bare number with no standard attached was not produced this way.
Comfort is the comparison column
The comfort figure is the one designed for choosing between bags, and it is the one this site compares. Because every bag quoting the standard has one, it is the column where a like-for-like comparison is actually possible.
Limit is a different reading of the same test
The limit figure is a second output of the same measurement, and it sits below the comfort figure by a margin that varies from bag to bag. That variation is itself informative: two bags with the same comfort figure and different limit figures are not identical bags.
What it is not is a second opinion you can substitute when the comfort figure is inconvenient. Comparing your candidate’s limit figure against another bag’s comfort figure produces a gap that exists only because you read different columns.
The lowest figure
A third figure sometimes appears at the bottom of the label. It is not a temperature to plan a night around, and it is not a rating this site ranks on. Its place on the label is a consequence of the standard producing it.
Where a maker publishes guidance on how the bag is intended to be used, that guidance — not the lowest number — is what to follow.
What none of the three include
The mat. A bag’s rating is measured with a mat under the mannequin, and the mat you actually own has its own R-value, measured to ASTM F3340 where the maker states it. Cold from underneath is the commonest reason a bag seems not to match its figures, and no figure on the bag describes it.
Fit is the other omission. An oversized bag is more air to warm; a bag too narrow at the shoulders is pushed open. Neither shows up in a temperature column.
How to read a label in thirty seconds
Find the standard. If it is not named, note that and move on to a bag that names one. Find the comfort figure and write it down. Note the limit figure as context rather than as an alternative. Ignore the lowest figure for the purposes of choosing.
Then read the fill weight, the length and shoulder girth, and the R-value of the mat you will use with it — and follow the manufacturer’s instructions for the bag you buy.
The manufacturer's own instructions come first, and this page does not stand in for them. Every weight, packed length, hydrostatic head, denier, micron rating, lumen figure and temperature on this page is reported from the manufacturer's own documentation, with its unit written as published, and named as such — none of it is measured here. A weight is reported as the maker counts it, and is never restated as another kind of weight, because the two are not the same figure. And the statement repeated rather than assumed: a comfort, limit or extreme temperature is a laboratory figure measured to EN 13537 or ISO 23537 on an instrumented manikin, not a promise about you. How cold a night you can spend is answered by the bag's own instructions, by what you sleep on and wear, and by conditions nothing here can know. Nothing on this site is advice about hypothermia.
What people ask before buying
Why do makers print the lowest figure at all?
Because the standard produces it, and a maker testing to EN 13537 or ISO 23537 publishes what the test yields. Its presence on a label is not a recommendation, and we do not rank bags on it.
Which figure should I compare between two bags?
The comfort figure, provided both makers name the same standard. Comparing one bag's comfort figure against another's limit figure produces a difference that is entirely an artefact of which column you read.
Last reviewed 6 September 2026