The standard answer to this question is body condition scoring, and it is a good answer. It takes thirty seconds, it needs no equipment, and every major veterinary body recommends it.
It is also less precise than it looks, harder for owners than anyone admits, and was never validated in the animal you are trying to use it on. All three of those things are worth knowing before you lean on it, and none of them means you should not use it.
How to actually do it
You do this by hand. The WSAVA Nutritional Assessment Guidelines are explicit that “palpation is required for accurately assessing BCS … especially in animals with medium to long hair coats.” Palpation means feeling with your fingers, and looking at a fluffy puppy only tells you about the coat.
Three checks:
- Ribs. Stand behind the puppy, rest both thumbs on the backbone, and spread your fingers down over the rib cage. At an ideal score the ribs are easy to feel under a light press (palpable, in the chart’s wording) under only a thin layer of fat. Push hard to find them and there is too much fat.
- Waist, from above. Stand over the puppy and look straight down. Behind the ribs the body should narrow, then widen again at the hips.
- Tuck, from the side. Crouch down. The line of the belly should slope upward from the ribs toward the back legs rather than running level. That slope is the abdominal tuck.
The WSAVA 9-point chart gives the descriptors verbatim:
| Score | Descriptor |
|---|---|
| 4 | Ribs easily palpable, minimal fat covering. Waist easily noted from above. Abdominal tuck evident. |
| 5 | Ribs palpable without excess fat covering. Waist observed behind ribs from above. Abdomen tucked up from side. |
| 6 | Ribs palpable with slight excess fat covering. Waist discernible from above but not prominent. |
| 7 | Ribs palpable with difficulty; heavy fat cover. Fat deposits over lumbar area and tail base. Waist absent or barely visible. |
| 8 | Ribs not palpable under very heavy fat cover, or only with significant pressure. Waist absent. No abdominal tuck. |
WSAVA’s target: “The goal for most pets is a BCS of 4 to 5 of 9.” BCS is body condition score.
Check which scale you are being handed. A 5-point system is also in wide use, on which 3 is ideal and 4 is overweight. Large prevalence studies (surveys of how common overweight condition is) have used it. A score of “4” means opposite things on the two scales.
What a point is worth in pounds
Purina Institute, which owns the system, states that “each Body Condition Score unit above 5 is associated with a 10 to 15% increase in body weight.”
Two honest caveats. The figure is given for points above 5 only. We found no source saying it works the same way below ideal, so do not turn a score of 3 into “15% underweight.”
And a rival figure of 5% per point does the rounds in the clinical press. Our best reading is that 10–15% is about body weight while 5% is about body fat, and that the two get mixed up. We could not confirm that from a primary source, though, so we would rather flag the clash than paper over it.
While we are here. A widely shared 2025 review states the 10–15% figure and hangs a citation on it. Follow the citation into its own reference list and you land on a paper about melanocortin receptor genetics in Beagles. That is how a number comes to feel settled when nobody has checked it.
How accurate is it, really
Not as accurate as its popularity suggests, and we would rather you knew.
Witzel and colleagues (2014) checked scoring against DEXA, the reference standard for measuring body composition (how much of a dog is fat and how much is lean) in overweight and obese dogs. The body condition score landed within 10% of the measured body fat in 6 of 47 dogs (13%). A body fat index method managed 25 of 47.
Christie and colleagues (2024) had four people score the same 43 working dogs, each on their own: a handler, a sports medicine resident, a board-certified specialist and a family vet. Interobserver agreement (how closely independent scorers match) came to R² = 0.62, and all four landed on the same exact score in 7 of 43 dogs.
The honest reading is that body condition score is good at direction and poor at absolutes. It will tell you whether a dog is drifting up or down over months. It will not tell you that your puppy is 14% over ideal.
We also could not check the original dog validation study (Laflamme, Canine Practice, 1997). The journal is defunct and not indexed, and no secondary source we read gives its sample size or its comparator, the thing scores were measured against. If you have seen “validated against DEXA at r = 0.87” attached to dogs, that figure comes from a study of cats. We are not going to repeat it here.
Owners get this wrong, and a chart does not fix it
This is the most uncomfortable finding in the literature, and it applies to everyone reading this, including us.
Eastland-Jones and colleagues (2014) asked 110 owners to assess their own dogs. 65% got it wrong, and of those, 92% underestimated. Then they gave the owners a body condition chart. 64% were still wrong, no improvement at all, while 77% believed the chart had helped them.
Yam and colleagues (2017) found much the same. Of 174 owners, 55% got their dog’s condition right from written descriptions and 51% from pictures, and 78% of the errors were underestimates. Only 48% could estimate their dog’s weight to within 10%.
What does work is being shown. Söder and colleagues (2023) found owners scoring their own dogs well below what professionals scored them. After a standard teaching session, owner scores were statistically indistinguishable from the professionals’, close enough that the numbers could not be told apart.
The practical point is not “don’t bother.” It is this: get a vet or nurse to put your hands on your own dog once and tell you what a 5 feels like. Reading a chart is measurably not the same thing.
Why a puppy is the hard case
Here is the gap. Body condition scoring has not been validated in growing dogs. Every validation study we could find used adults. Puppies are normally leaner than adults, and their shape changes week by week. So the descriptors are being used on an animal they were not written for.
The veterinary nutrition service at Tufts puts it directly. Holding a puppy at 4–5 out of 9 is critical, they say. But the score “is not sensitive enough in growing puppies to ensure they grow at the ideal rate.” That is expert clinical commentary rather than a study, and we are labelling it as such, but it is consistent with everything else here.
The tool that was built for growing dogs
What fills the gap is centile tracking, the same idea used on human babies: you plot the weight on a chart and watch which centile line it tracks.
Salt and colleagues (2017) built growth standards from Banfield practice records. They kept only dogs that stayed healthy and held ideal body condition through their first three years. The charts run from 12 weeks to 2 years, in five weight bands covering breeds up to 40 kg adult weight, with separate charts for males and females and nine centile lines on each.
The published usage rule, from WALTHAM’s own guide: “any sustained drop or rise through two or more centile spaces (crossing two centile lines) is unusual and should be carefully assessed.”
That threshold has been tested. Salt and colleagues (2020) found that fewer than 5% of healthy dogs cross more than two centile lines. But 68% of the dogs that were obese by three years old had crossed two or more lines upward during growth, and of the dogs that ended up underweight, 49% had crossed two or more lines downward.
The same authors are careful about their own tool: growth standard monitoring “is not perfect and should be considered alongside other methods of assessing weight status such as body condition scoring.” Neither replaces the other.
Two limits. The charts stop at 40 kg adult weight, which leaves out exactly the giant breeds where growth questions bite hardest, the same cut-off we apply in our own methodology.
And you will see the charts described as built from six million puppies. The paper says over six million young dogs. WALTHAM’s guide describes roughly 50,000 puppies contributing six million weight readings. Those are different quantities, we could not reconcile them, and so we are not quoting a headline number.
“But the breed standard says he should weigh…”
Breed weight ranges are a poor substitute. Smith and colleagues (2018) scored 140 purebred dogs and concluded that “published recommended, breed-specific body weight ranges are not a good predictor for an ideal BCS.”
That applies to our own tool too. A weight prediction tells you roughly where a dog is heading. It does not tell you whether the dog in front of you is carrying the right amount.
Why it is worth doing now rather than later
Montoya and colleagues (2025) looked at roughly 4.93 million dogs in US primary practice records. Overweight condition ran at 0.9% in early growth, 9.5% in late growth, and 24.4% in young adults. Their conclusion is the part that matters. Overweight or obese condition during growth persists into adulthood in most affected animals.
And it is not only a weight problem later. The same Labrador colony shows what extra weight during growth does to a large-breed puppy’s joints. Osteoarthritis (wear-and-tear joint disease) set in a median of six years later in the leaner dogs.
The window in which this is easy is short, and it is the one you are in.
This week
- Put your hands on your puppy and run the three checks. Aim for 4–5 out of 9.
- Weigh monthly to 6 months, then every 2–3 months. Plot the weights on a chart rather than just noting them down.
- Ask your vet to score your puppy alongside you once, so you calibrate your hands against a professional rather than a chart.
- Treat a steady climb across two centile lines as a prompt to adjust portions, not as a crisis.
If it is the amount in the bowl you are trying to set, our companion piece on how much to feed a growing puppy covers the sums and their limits. And if you need an adult weight estimate to plot against, the puppy size calculator will give you one.