If your dog is an Alaskan Malamute or a Siberian Husky, the answer is no and you just got five minutes back. (Rain is a different story for those two — that part is further down.)
For everyone else: the thing that decides this isn’t size, and it isn’t the thermometer. It’s whether your dog has an undercoat. That’s a thirty-second check you can do with your hand, and it’s more reliable than any breed list — including the one below.
The thirty-second test
Stand your dog up and part the fur at the shoulder, just behind the neck. Push your fingers straight down toward the skin.
Double coat. Your fingers meet resistance before they reach skin. Underneath the stiff outer guard hairs there’s a soft, dense, wool-like layer that springs back when you let go. In autumn you’ll come away with hair on your hand. That layer is the undercoat, and it is the insulation.
Single coat. Your fingers reach skin easily. There’s one layer of hair, it lies flat, and nothing springs back. It might be short, it might be long, it might be curly — none of that matters. There’s no second layer.
Do this before you trust any breed chart, because breed is only a shortcut. Individuals within a breed vary, and a mixed dog is anyone’s guess. The temperature side of this question is a separate calculation — this one is about your dog.
Three groups, not a sliding scale
Needs a coat
Short single coats
Dobermans, Boxers, Pit Bull types, Great Danes, Weimaraners, Vizslas, Greyhounds, Whippets.
Needs one — and gets missed
Plenty of hair, still single
Standard Poodles, Portuguese Water Dogs, most doodles, Soft Coated Wheatens, Kerry Blues.
Not a coat — a shell for wet days
Double coats
Huskies, Malamutes, Bernese, Newfoundlands, Australian Shepherds, most German Shepherds.
Labradors and Golden Retrievers sit between Group 2 and Group 3 and get their own note below.
Group 1 — short single coats
These are the dogs most people get right, because they look cold. Coat down to nothing, body fat in the single digits, and a lot of surface area for the mass.
Dobermans. Single short coat over a deep chest and very little fat, plus a high metabolism that burns through reserves fast. The short coat reads as “tough,” which is exactly backwards. A healthy adult Dobe is usually asking for a layer somewhere around 40°F, sooner if it’s wet.
Boxers. Muscle is not insulation. A Boxer carries almost no fat under a coat thin enough to see skin through, and the flat face means less efficient warming of the air it breathes. Same range as a Doberman.
Pit Bull types. Short, single, close-lying. Same story — a solid, muscular dog that reads as weatherproof and isn’t.
Great Danes. Big body, almost no coat. The size helps a little with heat retention, but not enough to cancel a coat that’s barely there. The real problem for Dane owners isn’t knowing they need something — it’s finding one that fits a dog that size. Most size charts stop before they start.
Weimaraners and Vizslas. The thinnest coats on this list, on lean athletic frames. If it’s below freezing, they need something.
Greyhounds and whippets are the extreme case for cold, and the hardest to fit. A sighthound’s chest-to-waist ratio is unlike any other breed’s, so measure before you buy anything from anyone — a pattern cut for a broad-chested dog will not sit right on one.
For everything else in this group, the question on a dry cold day is insulation and coverage — something that wraps the chest and belly, not just the back.
Group 2 — plenty of hair, still single
This is the group that gets missed, and it gets missed for an obvious reason: these dogs look warm.
Standard Poodles. A poodle’s coat is one continuous layer of curly hair with no undercoat underneath it. It looks like a lot of insulation. It isn’t. Do the shoulder test on a poodle and your fingers go straight to skin. A clipped poodle in January has less between it and the weather than a Labrador does.
Portuguese Water Dogs, Irish Water Spaniels, Barbets. Same construction, same answer.
Soft Coated Wheatens and Kerry Blues. Single-coated terriers on the edge of our size range — same test, same result.
Doodles. Depends entirely on which side won. See the mixes section below.
There’s a second thing about this group that the first thing hides. A curly single coat doesn’t just fail to insulate — it holds water, and it takes hours to dry. A soaked poodle is carrying the water and getting no insulation in return, which is the worst of both. So this is the one group where the honest answer is usually both: a warm layer for dry cold, and a shell for anything wet.
One caveat from a customer, in her words: “With standard poodle long legs the coat doesn’t go all the way down.” Poodles are built tall for their back length. If your dog is long in the leg, expect full-body coverage to stop above the ankle — and check the measurements rather than the breed.
Group 3 — the ones that don’t need one
German Shepherds — run the test on this one. Shepherds vary more than any other breed on this page. The classic stock coat carries a dense undercoat and is comfortable at 20°F with nothing on. But short-coated shepherds are common, plush and long-coats behave differently again, and a senior, a recently shaved dog or a genuinely hard winter can move any of them. This is the breed where the breed name tells you the least.
Australian Shepherds and Border Collies. Same answer.
Bernese, Newfoundlands, Saint Bernards. Built for cold, and warm enough without help. What these coats do badly is dry — see the Malamute note below.
Huskies and Malamutes. Not for cold — they work in temperatures far below anything you’ll walk in. Rain is a different question, and there the answer flips. A coat that dense holds water like a sponge: once it soaks through it stops insulating and starts pulling heat out, and it can stay damp for hours after you get home. A shell that keeps the undercoat dry does more for a Malamute than any amount of insulation will.
Labradors and Golden Retrievers. The middle ground. Technically single-coated, but dense, with an oily outer layer that sheds water reasonably well. Usually fine down to freezing. On long wet days a waterproof layer does more for them than a warm one.
Mixes, doodles and rescues
This is where breed charts stop being useful, and it’s a bigger group than most articles admit.
Doodles are a lottery. An F1 cross, an F1b backcross and a multigen doodle can have three completely different coats, and littermates often differ. Some inherit the poodle’s single curly coat and belong in Group 2. Some inherit enough retriever or Bernese to have a real undercoat and belong in Group 3. The dog’s pedigree won’t tell you. Your hand will.
Retriever, shepherd and husky mixes usually inherit the double coat, but “usually” is doing real work in that sentence — check.
Rescues of unknown background. The shoulder test is the only method you have, and it’s the only one you need.
One practical consequence: with a mixed dog you can’t size by breed either. Measure back length and chest girth and buy to the larger of the two, because the proportions won’t match any chart’s idea of a 70-pound dog.
Three things that change the answer
Wherever the groups above put your dog, any of these can move it.
This dog
- Age. Puppies and seniors regulate temperature less well. A ten-year-old Shepherd is not a three-year-old Shepherd.
- Arthritis and hip dysplasia. Common in big breeds, and cold makes stiff joints stiffer. For a lot of owners this — not the temperature — is the actual reason they buy a coat.
- Hypothyroidism. Underdiagnosed in large breeds, and it directly impairs temperature regulation. A dog that suddenly seems cold this year and didn’t last year is worth a vet conversation.
- Heart, kidney or endocrine disease, or anything that limits movement — a dog that can’t trot to stay warm gets cold faster.
- Recent surgery or a shaved patch. A clipped double-coated dog can take a year or more to grow the undercoat back. For that year, it’s in a different group.
- Body condition. Fat is insulation. A dog that has lost weight this year has lost some of it.
This particular walk
- Pace. A dog trotting is generating its own heat. A dog stopping to sniff every three feet isn’t. The same dog at the same temperature can need a layer on the slow walk and not on the run.
- Standing time. Waiting outside the vet, on the sideline of a kid’s game, between runs at a training class, at a trailhead while you sort out the car. This is where dogs get cold, not on the move.
- Duration. A temperature that’s fine for ten minutes isn’t fine for an hour.
- Wind. It strips the warm air sitting in the coat. On an exposed field, 30°F behaves like something colder.
- The ground. Wet grass, slush and snow pull heat out of the belly and paws faster than the air does — which is why belly coverage matters more than back thickness.
This winter
- Acclimation. A dog that’s been out every day through a gradual autumn handles 25°F far better than the same dog would in a sudden October cold snap. The Tufts weather scale gives acclimation a full point.
- A recent move. A dog that spent last winter in Phoenix is not acclimated to this one.
- A first winter. Puppies and newly adopted dogs have no baseline, and neither do you.
- A warm house. A dog that lives at 72°F feels 35°F more sharply than one that doesn’t.
Why the fit is harder on this side
Two things go wrong more often with big dogs than small ones.
Deep chests. Dobermans, Danes, Weimaraners and most of Group 1 have a large gap between chest girth and waist. A coat picked on back length alone will almost always bind at the shoulder — and a coat that restricts the shoulder is one your dog will fight you about every time.
Giant breeds. Most size charts simply stop. If your dog is past 100 pounds you already know this.
Our own sizing runs generous through the chest for exactly that reason, which has a consequence people don’t expect: if your dog is lean and short-coated, size down rather than up. The common mistake is ordering too big, not too small. The size guide shows where to put the tape — and the rule is to buy to the larger of back length and chest girth, not to average them.
If the test put your dog in Group 1 or 2, the calendar matters more than the thermometer: they’ll want something from about the first frost. If it put them in Group 3, the honest answer is to skip the winter coat and spend the money on a shell for wet days instead.