A puppy growth chart is a laminated sheet on an exam-room wall with nine wavy lines printed on it, and it is a monitoring instrument rather than a crystal ball. Your vet dots your puppy's weight onto it, then dots it again at the next visit, watching whether the second dot sits on the same line as the first. That is the whole trick. Not how big will he get — is he still on the track he started on.
I'm Dana Whitcomb, a certified veterinary technician twelve years in, and I have put a lot of biro dots on those sheets. This page is that sheet, in numbers. If what you want is a prediction instead — how big is this puppy going to end up — that job belongs to our puppy weight calculator, which runs these same curves backwards. Different question, different tool.
The puppy growth chart, in numbers: percentage of adult weight by week
Here is what the lines actually say. At any age, a healthy puppy of a given size class and sex has reached a fairly predictable percentage of its adult weight. Find your column, read down.
| Size class | 12 wk | 16 wk | 20 wk | 24 wk | 26 wk | 30 wk | 36 wk | 44 wk | 52 wk |
|---|---|---|---|---|---|---|---|---|---|
| Toy (under 14 lb) | 41% | 54% | 65% | 73% | 77% | 83% | 89% | 95% | 99% |
| Small (14–20 lb) | 38% | 52% | 64% | 73% | 77% | 83% | 89% | 94% | 98% |
| Medium (20–33 lb) | 32% | 46% | 58% | 68% | 72% | 79% | 86% | 92% | 96% |
| Large (33–66 lb) | 29% | 42% | 55% | 66% | 71% | 80% | 88% | 94% | 97% |
| Extra large (66–88 lb) | — | 42% | 55% | 67% | 72% | 80% | 89% | 95% | 98% |
| Giant (over 88 lb) — no standard exists | — | 38% | 50% | 60% | 65% | 73% | 82% | 89% | 94% |
Digitized by Paw Charts from the published WALTHAM growth charts — paw-charts.com
| Size class | 12 wk | 16 wk | 20 wk | 24 wk | 26 wk | 30 wk | 36 wk | 44 wk | 52 wk |
|---|---|---|---|---|---|---|---|---|---|
| Toy (under 14 lb) | — | 57% | 67% | 75% | 78% | 83% | 89% | 95% | 99% |
| Small (14–20 lb) | — | 53% | 64% | 73% | 77% | 82% | 89% | 94% | 97% |
| Medium (20–33 lb) | 35% | 48% | 58% | 67% | 71% | 77% | 84% | 90% | 95% |
| Large (33–66 lb) | 31% | 44% | 57% | 68% | 72% | 80% | 87% | 93% | 96% |
| Extra large (66–88 lb) | 30% | 44% | 56% | 67% | 72% | 79% | 87% | 93% | 96% |
| Giant (over 88 lb) — no standard exists | — | 40% | 51% | 61% | 66% | 72% | 80% | 87% | 92% |
Digitized by Paw Charts from the published WALTHAM growth charts — paw-charts.com
The dashes aren't laziness — those curves have no data that early, so we print nothing rather than a guess. And look at the shape of it: a toy dog is 41% grown at 12 weeks and essentially finished by its first birthday, while the bottom row is still adding weight past 12 months. Same species. Two entirely different timetables.
Puppy growth stages, as the numbers see them
Stages in this subject are usually described in adjectives. Here they are in figures, using the large-breed male curve as the spine, with the skeletal timings from von Pfeil and DeCamp's 2009 veterinary review of the growth plate.
| Age | Large male, % of adult | What is going on |
|---|---|---|
| Under 12 weeks | — | No growth standard exists. The charts begin at 12 weeks and nothing published covers before it. |
| 12 weeks | 29% | The first dot you can legitimately plot. |
| 12–26 weeks | 29% → 71% | "In dogs, major growth occurs between 3 and 6 months of age." Also when most growth plates start closing — the range is 4 to 12 months by site and breed. |
| 36 weeks (~9 months) | 88% | "Most dogs achieve 90% of their adult size by the end of 9 months." The curve and the quote agree. |
| 52 weeks (12 months) | 97% | Nearly done by weight. The AKC puts a large-breed dog at 85% here — sources genuinely disagree. |
| 15–18 months | ~100% | Only giant breeds are still meaningfully changing, and even that is the authors' stated clinical impression rather than measured data. |
Digitized by Paw Charts from the published WALTHAM growth charts — paw-charts.com
Looking for puppy growth week by week pictures?
Then I should be straight with you: we don't have any. No photographs, and I'm not going to pad the page with stock spaniels to catch the search. What we have is the numbers behind what those pictures would show — the tables above are the week-by-week record, written in percentages rather than fur. Honestly, it's the more useful artifact. A photo of somebody else's 16-week Labrador tells you about that Labrador. The curve tells you what yours should weigh, which you can check on a scale in your own kitchen.
Where these numbers come from, and what they are not
The gold standard is a 2017 PLOS ONE paper by Salt and colleagues, which built growth standards using GAMLSS modelling with a Box-Cox Power Exponential distribution, fitted to Banfield Pet Hospital records from 1994 to 2013 across roughly 900 US veterinary hospitals — about 1.06 × 10⁵ data points across 3.3 × 10⁴ dogs, screened to animals in ideal body condition. It is the best thing in this subject and it isn't close.
But the paper publishes its standards as graphs only. There are no numeric percentile tables — not in the paper, not in its supplement, not in its public dataset. Nobody can look these numbers up, which is a strange thing to be true of the most important growth data in the field, and it is why every chart online is either a picture or a fabrication.
So we made the numbers. We took the published WALTHAM chart PDFs — male and female, all five categories — extracted the vector coordinates of the centile curves, and calibrated against the printed axes. The calibration self-validates, which is the only reason I'll publish it: every x gridline resolves to exactly 1-week spacing, every y gridline to round kilogram steps, and all nine curves on all ten charts begin at precisely 12.0 weeks — independently reproducing the paper's own stated starting point, which we had not told the extraction to look for.
These are our measurements of somebody else's printed curve, not WALTHAM's published figures — there are no published figures. Expect roughly ±1–2% reading error on every percentage here. And one caveat nobody in this industry states: the WALTHAM charts are a clinical monitoring tool. Using them to predict an adult weight is our inference, not their purpose. Sources and dates on how we calculate.
Puppy growth chart by breed size: why five classes, and why nobody agrees
The five WALTHAM categories weren't picked by convention or rounded to tidy numbers. They came out of visual assessment and hierarchical cluster analysis of breed-specific growth curves — dogs grouped by how they actually grow, with the boundaries falling where they fell. That's why they sit in awkward metric brackets rather than clean imperial ones.
The AKC uses four classes instead, by imperial weight, with no toy class at all. Its Small bracket runs 0–20 lb, swallowing all of WALTHAM's first category, all of its second, and part of its third. Its Large bracket straddles one category WALTHAM modelled and one it refused to. These two sources have never described the same dog the same way.
The disagreement, quantified — because it is large
At six months the AKC says a large-breed dog is 60% grown. Our WALTHAM digitization says 71–72%. At twelve months, AKC says 85%; the curves say 97–98%. That is not rounding. On a dog headed for 70 lb, those sources disagree by about eight pounds at six months — in the direction that decides which crate you buy and whether you panic about the food bill.
| AKC size class | 6 mo | 9 mo | 12 mo | 15 mo | 18 mo |
|---|---|---|---|---|---|
| Small (0–20 lb) | 75% | 90% | 100% | 100% | 100% |
| Medium (21–50 lb) | 66% | 85% | 95% | 100% | 100% |
| Large (51–100 lb) | 60% | 75% | 85% | 95% | 100% |
| X-Large (100+ lb) | 50% | 65% | 80% | 90% | 100% |
Source: the American Kennel Club's published growth table. Reproduced for comparison, not endorsement.
We favour the curves, and here is the reason rather than the assertion: the AKC's table carries no citation and names no data source, while the WALTHAM figures come from a peer-reviewed model fitted to roughly 33,000 dogs. That's the whole argument. We publish both anyway, because picking a side quietly and pretending there was never a conflict is how this niche got into the state it's in.
There is no growth standard for giant breeds. None.
The researchers tried. Growth among breeds over 40 kg — 88 lb — was too inconsistent to combine into one curve, so they published five categories and left the sixth blank. Royal Canin's own veterinary teaching material says the same. Which means every Great Dane, Mastiff and Newfoundland chart on the internet is an extrapolation with no evidence underneath it, including the giant row in our tables above. We print it because people need something. We label it because they deserve to know what they're getting.
When do puppies stop growing?
By size, and the spread is wider than most people expect. Our digitization has the median dog reaching 99–100% of its chart-end weight at about 52 weeks for toy and small breeds, 62 weeks for medium, 67–68 weeks for large and extra large, and roughly 78 weeks for giants — that last figure carrying the caveat above.
The AKC disagrees again, and this time our data is the one saying slow down: it puts small breeds finishing at six to eight months. The toy curve has those dogs at 83% at 30 weeks, which is about seven months, and not at 99% until roughly a year. Both figures are on this page; I'm not averaging them into a number that belongs to nobody. A veterinary textbook chapter lands between the two — toy, small and medium dogs at adult size by 9 to 10 months, giant breeds not reaching 99% until 11 to 15.
Growth plates, and why the big dogs finish late
Weight is what you can measure at home, but the growth plate sets the timetable — a band of cartilage near the end of each long bone that lengthens it and then ossifies and stops. Once it closes, that bone is done. von Pfeil and DeCamp put most closures between 4 and 12 months depending on the anatomic site and breed, and note that closure comes earlier in smaller animals. For giants they add that some plates may not close until 15 to 18 months — and they explicitly flag that as clinical impression rather than measured data, which is a careful thing to do and worth repeating. Anyone publishing a tidy closure-age-by-size table invented it.
Puppy growth spurts are mostly a story we tell
The curves are smooth. Real puppies aren't, and "he had a growth spurt last week" is the single most common thing people bring me from a chart. Maybe. But the curve you're comparing against is a population average built from tens of thousands of dogs, and averaging smooths everything. Individual puppies grow in fits and starts around that trend; a week of visible catching-up is normal variation rather than an event.
It matters because a spurt and a drop are the same phenomenon read in two moods. If a weigh-in puts your puppy 8% above where you expected, that's not a spurt — that's one measurement and one expectation, and the honest read is that you need a third dot.
When to actually worry — and when not to
Here is the most clinically useful thing on this page, and almost nobody says it: being off the average is not the signal. Crossing lines is.
The same group measured it in 2020. In healthy dogs, 42% crossed more than one centile line during growth — nearly half of perfectly normal puppies drift off the line they started on — but fewer than 5% crossed more than two. In dogs with abnormal body condition the picture inverts: 68% of obese dogs crossed at least two centile lines upward, 49% of underweight dogs crossed at least two downward, and 54% of dogs with growth-accelerating disease crossed at least two.
Read that both ways. A puppy sitting on a low line and staying on it is usually just a small puppy — if you take one thing from this page, take that one. A puppy moving across lines, consistently, in one direction, is the pattern that earns a phone call. The number that matters isn't the gap between your dot and the middle line. It's the gap between your dot and your own last one.
Three measurements, two weeks apart — and the thing our tool cannot do
WALTHAM's own guidance is three serial measurements, roughly two weeks apart, after 12 weeks, before you trust a track at all. That's their instrument and their instruction. No one-shot calculator does it, and I include ours: on a single visit the tool above has one dot and cannot tell a trend from a fluke, so it doesn't pretend to. Weigh him again in a fortnight and come back. Three points beat one, and it costs you nothing but a note on the fridge.
Reading the chart when the breed is a guess
Mixed breed puppy growth chart
There isn't a separate one, and there doesn't need to be. A mixed breed puppy growth chart is just the curve for the adult size you expect — pick the class, plot the dot. The hard part isn't the chart; it's knowing which class to pick.
Parents' weights beat everything else available to you. Body size in dogs is oligogenic — it runs on a handful of large-effect genes rather than thousands of tiny ones, which is unusual in biology and exactly why "how big were the parents" is such a good question. Six loci explain 64.3% of size variance across breeds up to 41 kg, and a single IGF1 haplotype is common to every small breed and nearly absent from the giants. Know both parents? Average them. Don't? A DNA test beats a guess, and a guess beats nothing.
Lab puppy growth chart, German Shepherd puppy growth chart — same curve
People search these separately, so let me collapse them. If you want a lab puppy growth chart or a german shepherd puppy growth chart, you want the same row: both breeds sit in WALTHAM's category V, 30–40 kg or 66–88 lb, which appears in our tables as extra large. The paper names them together — German Shepherd Dog, Golden Retriever, Labrador Retriever, American Bulldog — because they cluster on how they grow, not because somebody sorted them alphabetically. So an extra-large male is 42% of his adult weight at 16 weeks, 72% at six months, 98% at a year. There is no separate Labrador science.
Wendell, at 14 weeks
My own dog, since he's the only patient story I'm entitled to tell. Wendell is a 62 lb Labrador-hound mix, neutered, seven years old, asleep against my chair as usual. At 14 weeks he weighed 22 lb. A large-breed male at 14 weeks is 35.5% of his adult weight, and 22 ÷ 0.355 gives 62 lb. He is 62 lb. That is the arithmetic our puppy weight calculator runs, and I want to be careful about it: that's a curve landing dead centre and it won't always do that. What I'd actually have wanted at the time was three dots, not one.
A growth chart is not a prediction tool, and it was never built to be one. WALTHAM built these curves to monitor — to catch a puppy sliding off its own track early enough to matter. The value of the instrument sits in the second and third measurement. The first one is just where you start.
Then the industry repurposed it, and I'm not pointing at anybody else here — we do it too, on our own homepage, because "how big will my puppy get" is what people type. Turning a monitoring chart into a one-shot predictor is a defensible inference and the arithmetic works. But it isn't what the instrument is for, and it quietly drops the part that made it valuable. A chart used once is a chart used wrong. Weigh him again in two weeks.
What else is on the fridge
Once the growth question settles, the next one is the food bowl — and the honest answer there is that you can't feed a puppy bigger, only faster, which in a large-breed puppy is a problem rather than an achievement. The dog food calculator has a puppy mode that works from your own bag's calorie number rather than the generic chart on the back of it. Cats grow on a completely different clock; that one has its own reference in the kitten growth chart. And when this stops being a puppy question at all, the dog age calculator shows how differently these size classes age — the same gap that makes giant breeds grow slowly makes them get old fast. Every percentage on this page, with its source and the date we last checked it, is listed on how we calculate.
Frequently Asked Questions
When do puppies stop growing?
It depends on size, and the spread is wider than most people expect. Our digitization of the WALTHAM curves has the median dog reaching 99 to 100% of its adult weight at about 52 weeks for toy and small breeds, 62 weeks for medium, 67 to 68 weeks for large and extra large, and roughly 78 weeks for giants. Sources disagree and we show both: the AKC says small breeds finish at 6 to 8 months, which the curve data contradicts — toy dogs are still at 83% around seven months. A veterinary textbook chapter lands between them, putting toy, small and medium dogs at adult size by 9 to 10 months and giant breeds not reaching 99% until 11 to 15 months.
My puppy is below the line on the growth chart. Should I worry?
Not by itself, no. In the published follow-up study, 42% of healthy puppies drifted across more than one centile line during growth and turned out completely normal. A puppy tracking along a low line is usually just a small puppy. What matters is direction and consistency: among dogs in abnormal body condition, 68% of obese dogs crossed two or more centile lines upward and 49% of underweight dogs crossed two or more downward. So the signal is crossing lines repeatedly in one direction, not sitting below the average. Compare today's dot to your own previous dots, not to the middle of the chart.
How do I read a puppy growth chart?
Plot weight against age in weeks, find the line your puppy sits nearest, and follow it. The chart is not asking whether your puppy is average — it is asking whether your puppy is still where your puppy was. Read the tables on this page by size class and sex, since the published charts genuinely differ between males and females and we keep them separate rather than averaging. One rule from WALTHAM that no calculator can do for you on a single visit: take three measurements, about two weeks apart, after 12 weeks, before you trust a track. One dot is a data point. Three are a trend.
Is there a growth chart for a mixed breed puppy?
It is the same chart — you just pick the size class yourself. There is no separate mixed breed growth standard and there does not need to be, because the curves are organised by adult size rather than by breed. The useful input is the parents' weights if anyone knows them; average the two and use that as the expected adult size. Dog size is oligogenic, meaning it runs on a small number of large-effect genes — six loci explain 64.3% of size variance in breeds up to 41 kg — which is exactly why parent weights predict so well. Failing that, a DNA test gives you a mix you can size-class from.
Do puppies have growth spurts?
Usually not in the way people mean. The published curves are smooth because they are population averages built from tens of thousands of dogs, and averaging smooths everything out. Individual puppies do grow in fits and starts around that trend, so a week where yours seems to catch up is normal variation rather than an event. The practical problem with the word spurt is that it turns two measurements into a story. If a weigh-in puts your puppy above or below where you expected, that is one dot and one expectation. Get a third before you decide what it means.
Why is there no growth chart for giant breeds?
Because nobody has built one. The researchers behind these growth standards tried and could not: growth among breeds over 40 kg — 88 lb — was too inconsistent to combine into a single curve, so they published five size categories and left the sixth blank. Royal Canin's veterinary teaching material says the same. That means every Great Dane, Mastiff and Newfoundland growth prediction online is an extrapolation with no evidence base underneath it, including the giant row in our own tables. We publish it because people need something to work with, and we label it because they should know what they are getting.