Warmth in an insulated jacket comes from loft: the thickness of trapped still air the garment holds around you. Fill weight and construction determine that loft. Fill power only describes how much loft a given mass of down produces, so it is a weight and packability specification, not a warmth specification. Two jackets with the same fill weight and different fill powers are close in warmth and far apart on the scales.
Where the warmth actually comes from#
Insulation works by immobilising air. Still air conducts heat at about 0.026 watts per metre kelvin, which is poor, and that poorness is the product you are buying. Down clusters and synthetic fibre webs both exist to hold air in place and stop it circulating. The thicker the layer of trapped air, the higher the thermal resistance, close to proportionally.
So the chain runs: fill weight and construction produce loft, loft produces resistance, resistance produces warmth. Every marketing claim that skips a step should be read carefully. A jacket described only by its fill power has told you nothing about how warm it is, because you can put 60 g or 200 g of the same down into the same shell.
Loft is also what you lose first. Dirt, body oils and compression all reduce it, and a jacket that has lost a third of its loft has lost close to a third of its insulation regardless of what the label says. That is the entire argument of washing down gear and of storing it uncompressed, covered in gear storage.
What fill power tells you, and what it costs#
Fill power is the volume in cubic inches that one ounce of down occupies under a light standard weight after conditioning, measured under the IDFB Testing Regulations. Commercial down runs roughly 550 to 900. Higher numbers mean each gram lofts further.
Put arithmetic on the difference. One hundred grams is 3.53 ounces, so 100 g of 850 fill power down lofts to about 3,000 cubic inches (3.53 x 850). To reach the same 3,000 cubic inches with 650 fill power down you need 3,000 divided by 650, which is 4.62 ounces, or about 131 g. The 850 fill jacket is therefore about 31 g lighter for the same warmth, and it compresses smaller. It is not warmer.
That 31 g is a real benefit and a modest one. Where high fill power earns its price is at the extremes: on an expedition parka carrying 300 g of fill, the same ratio saves nearer 95 g and a noticeable amount of pack volume. On a 60 g summer sweater it saves under 20 g.
Two cautions. First, fill power is measured after conditioning, and more than one conditioning method is in use, so figures from different regions are not always on the same scale; a number without a stated method deserves mild scepticism. Second, fill power says nothing about the down to feather ratio, which EN 12934 defines and which labels express as figures such as 90/10. The choice between down and synthetic fill is a separate question, worked through in down against synthetic insulation.
Static or active: the framework that decides everything else#
Most confusion about insulated jackets dissolves once you sort them into two categories with different design goals.
A static piece is built to be put on when you stop. It maximises loft per gram, uses a tightly woven low permeability shell to block wind, and is cut generously to fit over everything else. It is not designed to be walked in, and walking in one soaks it.
An active piece is built to be worn while moving. It uses lower loft insulation, often a synthetic designed to breathe, in a shell with deliberate air permeability so vapour and warm air can escape. It is less warm at rest by design, and it stays usable at output levels where a down piece would be a sauna.
| Use case | Temperature band | Type | Typical fill | Shell |
|---|---|---|---|---|
| Fast hiking and climbing in the cold | minus 5 to 5 degrees C | Active | Light synthetic, breathable | 5 to 20 CFM, air permeable |
| Cool evenings, three season camp | 0 to 10 degrees C | Static | 60 to 120 g down | Under 1 CFM, windproof |
| Winter camp layer, long stops | minus 15 to 0 degrees C | Static | 150 to 250 g down | Under 1 CFM, box baffles, hood |
| Damp maritime cold, multi day | 0 to 8 degrees C | Static | Mid weight synthetic | Under 1 CFM, water resistant face |
| Standing about in deep cold | below minus 20 degrees C | Static | Above 250 g down | Under 1 CFM, full hood and long cut |
The layering guide works out why the gap is so wide: modelled clothing insulation at 0 degrees C is around 1.2 clo while climbing and around 3.6 clo standing in camp. Those are two garments, not one.
Shell air permeability is the specification nobody reads#
Air permeability is measured in cubic feet of air per minute passing a square foot of fabric at a set pressure, under ASTM D737 or ISO 9237. Manufacturers who publish it typically quote under 1 CFM for a tightly woven windproof shell, roughly 5 to 20 CFM for the air permeable fabrics used on active insulation, and higher still for very open fabrics.
This one number, more than the insulation inside, decides how a jacket behaves. A near zero CFM shell holds warm air in place, which is exactly right when you are still and exactly wrong when you are producing 400 watts, because the vapour has nowhere to go and condenses in the fill. A 10 CFM shell lets vapour and some heat escape continuously, so the jacket runs cooler at rest and stays dry while you work.
Notice the consequence. Putting a windproof down sweater under a waterproof shell for a climb creates two vapour barriers in series, and the sweat has to condense somewhere. See rain jackets and breathability for why the outer half of that sandwich stops helping in warm humid conditions.
Baffles: where the warmth quietly leaks out#
Sewn-through construction stitches the outer and inner fabrics directly together, creating chambers. It is light, cheap and simple, and at every stitch line the insulation is pinched to nothing, leaving a cold stripe.
Model the cost. Take a jacket with 2 clo over most of its area, and assume the seams pinch 8 percent of the surface down to 0.5 clo. Heat flow scales with area divided by resistance, so the seamed jacket loses (0.92 / 2) + (0.08 / 0.5) = 0.62 units against 0.50 for a uniform 2 clo garment. That is about 24 percent more heat loss for the same fill weight. The assumptions are ours, but the direction is not in doubt.
Box wall construction adds vertical fabric walls between the layers so the insulation keeps its full thickness across the seam. It costs weight in fabric and stitching, and it is why serious cold weather pieces use it. The same logic governs sleeping bags, which is why a bag rating can differ several degrees between two bags with identical fill weight.
Sewn-through is the right choice for a light sweater used above freezing, where the weight saving matters more than the seam loss. Below about minus 10 degrees C for static use, box baffles earn their weight.
Hoods, collars and sizing over your layers#
A hood is the highest return per gram in this category. The head is roughly 7 to 10 percent of body surface area, it is well supplied with blood, and it is normally uncovered, so covering it removes a disproportionate share of your total heat loss for 60 to 100 g of jacket. A hood that fits over a helmet or a beanie, has a single hand adjustment and closes across the chin is worth more than 30 g of extra fill.
Sizing is where good jackets are ruined. A static piece must close comfortably over your moving layers, because that is precisely when you will wear it. But excess volume is not free either: air that can circulate inside the jacket carries heat away by convection, so a jacket two sizes too big loses some of what it gains. The target is enough room to compress nothing and no more.
Check the hem length as well. A jacket that sits above the harness or hip belt line exposes the lower back, which is a large area to leave uninsulated while sitting.
Check this before you buy#
Four checks you can run in a shop or at home, none of which need a lab.
- Read the fill weight, not the fill power. If a brand does not publish fill weight in grams, that is information in itself. Compare like for like: 100 g of fill in one jacket against 100 g in another.
- Wear your actual layers. Put on the base layer and mid layer you hike in, then try the jacket. Zip it fully, raise both arms, and pull the hood up over a hat. If any of that is a struggle, the size is wrong.
- Pinch the seams. Hold a baffle seam between finger and thumb. If it goes flat, it is sewn-through. Decide whether that matters for the temperatures you actually use.
- Test the loft after storage. Take the jacket out of its stuff sack and time how long it takes to recover full thickness. Down should regain most of its loft within twenty to thirty minutes of shaking. A piece that stays flat has either been compressed too long or needs washing.
Common mistakes#
Buying for the coldest ten minutes of your year. This is the big one. Most people choose a jacket imagining the coldest moment they can remember, then wear that jacket in conditions that occur on a handful of days. The result is a garment too warm to move in and too bulky to carry, so it stays at home and they end up cold in the merely chilly conditions they actually meet. Choose for the temperature band you are in most often, and treat deep cold as a separate problem.
Assuming higher fill power is warmer. It is lighter for the same warmth. The arithmetic above puts the gain at about 31 g per 100 g of fill between 650 and 850.
Hiking in a static down jacket. It has a windproof shell for a reason, and that reason is standing still. Moving in one drives sweat into the fill, where it destroys loft.
Storing it compressed. Months in a stuff sack costs permanent loft. Store insulated pieces hanging or loose in a large bag.
Ignoring the weight it saves elsewhere. A properly warm camp layer often lets you carry a lighter sleeping bag or fewer hedged mid layers, which is a base weight question rather than a clothing one.
Frequently asked questions#
How much down fill do I need in a jacket?#
Match fill weight to use rather than to a temperature promise. Around 60 to 120 g of down suits a light sweater for cool evenings and three season camp, and 150 to 250 g suits a warm parka intended for long stops in winter conditions. Below about minus 20 degrees C, static use calls for more than 250 g plus a hood and a long cut.
Is 800 fill power warmer than 650?#
No, it is lighter for the same warmth. Fill power measures the loft that one ounce of down produces, so higher fill power means less mass for the same thickness. One hundred grams of 850 fill lofts about as much as 131 g of 650 fill, a saving near 31 g. If both jackets contain 100 g, the 850 one is slightly warmer only because it lofts more.
What is active insulation and do I need it?#
Active insulation is designed to be worn while moving. It combines lower loft synthetic fill with an air permeable shell, typically quoted around 5 to 20 CFM, so vapour escapes instead of condensing. It is genuinely useful if you work hard in cold conditions and hate stopping to change layers. If your cold time is mostly spent standing still, a static piece is better value.
Down or synthetic for a camp jacket?#
Down gives more warmth per gram and packs smaller; synthetic keeps more of its loft when damp and dries faster. For dry cold and weight sensitive trips, down. For sustained damp, multi day trips with no chance to dry gear, or hard wearing use, synthetic. The trade is worked through in detail in the down against synthetic comparison.
Are sewn-through baffles bad?#
Not bad, just limited. They save weight and cost, and they leak heat along every seam line. For a summer sweater used above freezing that is a fair trade. For static use in serious cold, box wall construction preserves loft across the seams and is worth its extra fabric weight.
Should an insulated jacket have a hood?#
Almost always yes for a static layer. A hood costs 60 to 100 g and covers an area with high heat loss that is otherwise bare. It also lets you extend the useful range of the jacket without carrying a separate item. The exception is a jacket intended purely to be worn under a shell with its own hood.
What size should I buy if I wear it over other layers?#
Size so that it closes without compressing what is underneath, and no larger. Compressing your mid layer flattens its loft, and excess space inside allows air to circulate and carry heat away. Test with the layers you actually hike in rather than in a shop T-shirt.
Can I wear a down jacket in the rain?#
Only briefly, and only under a shell for anything sustained. Water repellent treated down resists a light shower and still loses loft once genuinely wet, and wet down takes a long time to recover. In persistent rain, a synthetic insulated piece under a waterproof shell is the more reliable arrangement.
Standards, sources and further reading
- IDFB Testing Regulations, Part 10, Fill Power, International Down and Feather Bureau. Defines the loft measurement behind fill power figures, including the conditioning methods that make numbers differ.
- EN 12934:1999, Feather and down: composition of processed feather and down filling material, CEN. Defines the down to feather ratios quoted on labels, such as 90/10.
- ASTM D737-18, Standard Test Method for Air Permeability of Textile Fabrics, ASTM International. The source of the cubic feet per minute figure quoted for shell fabrics. ISO 9237 is the equivalent method.
- ISO 9920:2007, Ergonomics of the thermal environment: estimation of thermal insulation and water vapour resistance of a clothing ensemble, ISO. Published clo values for insulated garments.
- ISO 11092:2014, Textiles: measurement of thermal and water-vapour resistance under steady-state conditions, ISO. The method behind evaporative resistance figures for insulated and shell fabrics.
How this page is made. Every number here is either a published standard, a physical constant, or arithmetic we show in full so you can check it. Read our evaluation method and editorial standards, or tell us we got something wrong.
Last reviewed and updated 9 September 2026.