Skip to content
PawCurve

Puppy Growth Chart

What puppies of each adult size actually weigh, week by week — computed from 8.5 million real veterinary weight records, and covering the early weeks the published charts leave out.

Sex
Puppy growth chart showing weight percentile curves by age

The same chart as numbers

Find your puppy's age in the left column. The columns are percentiles — the 50th is the middle puppy, and half of all healthy puppies sit outside the 25th–75th range, which is normal.

Age3rd25th50th (middle)75th97th

Why there isn't one puppy growth chart

People search for "the" puppy growth chart, and the honest answer is that a single chart is a category error. Puppies of different adult sizes don't grow on the same schedule — not just to different heights, but at genuinely different rates. We measured it:

Adult size% of adult weight at 8 weeksat 6 monthsFinishes growing
Toy25%70%~12 months
Small25%75%~12 months
Small–medium21%70%~12 months
Medium16%69%~15 months
Large16%67%~18 months
Giant12%62%~24 months

Look at the first column. A toy puppy at 8 weeks is already a quarter of its final weight; a giant-breed puppy is barely an eighth. That's a two-fold difference in how far along they are at the same age — which is why a chart that merges them describes no real dog, and why every "multiply your puppy's weight by X" rule falls apart across sizes.

The early weeks other charts skip

Here's something we didn't expect to find. The growth charts most sites cite start at 12 weeks. Ours start at 6.

That isn't because we have better data — it's the same data. When we downloaded the raw dataset, there were 741,559 weight measurements at 8–10 weeks and 293,948 at 6–8 weeks. The 12-week floor was a modeling choice by the paper's authors, not a limit of the records.

Which matters, because 8 weeks is when most people get their puppy — and it's precisely the moment they go looking for a growth chart. The published charts answer every week except the one being asked about. So we computed the curves down to week 6 and said so on the methodology page, including the fact that this is our extension rather than something the original paper endorses.

What a percentile is, and what it isn't

A percentile is a description, not a score. If your puppy is on the 25th percentile, three out of four puppies its age and size are heavier — and that is completely fine. Someone has to be on the 3rd. There is no prize for the 90th.

What the chart is actually good for is the trend. A puppy that tracks its own curve is doing what puppies do. A puppy that drifts across curves — 60th percentile in March, 20th in May — is worth a vet's attention, even though both of those numbers are "normal" in isolation. That's the thing a single weigh-in can't tell you and a chart can.

Two practical notes. Weigh at the same time of day, ideally before a meal — a puppy that just ate can read several percent heavier, which at 9 weeks is a real chunk of the number. And if you're unsure of the adult size class, try the chart at two sizes; a Great Dane puppy plotted on a medium chart looks alarming and is perfectly healthy.

Frequently asked questions

What should my puppy weigh at 8 weeks?
It depends entirely on the size the puppy is growing into, which is why one chart can't answer it. At 8 weeks the middle (50th percentile) puppy weighs roughly 2.6 lb if it's a toy breed, 10 lb if medium, and 13 lb if giant. Pick your adult size class in the chart above for the exact figures. Worth knowing: most published growth charts don't cover 8 weeks at all — they start at 12. Ours goes down to 6.
How do I read a puppy growth chart?
Find your puppy's age along the bottom, then see which curve its weight lands on. That's its percentile. A puppy on the 75th percentile is heavier than 75% of puppies its age and size class. Percentile is not a grade — someone has to be on the 3rd and someone on the 97th, and both can be perfectly healthy. What matters is whether your puppy stays roughly on its own curve as it grows.
My puppy is off the chart. Is that bad?
Not by itself. About 6% of perfectly healthy puppies sit outside the 3rd–97th band, because that's what those percentiles mean. Two things matter more than the number: first, whether you've picked the right adult size class — a Great Dane puppy plotted on a medium chart will look alarming and be fine. Second, the trend: a puppy that has been tracking the 60th percentile and slides to the 20th is worth a vet visit even though both are "normal" numbers.
Why does your chart go below 12 weeks when others don't?
Because the data does, and we went and looked. The widely-cited published growth charts start at 12 weeks — that was the authors' modeling choice, not a limit of the underlying records. When we downloaded the raw dataset we found 741,559 weight measurements at 8–10 weeks and 293,948 at 6–8 weeks. Since "what should my 8-week-old weigh" is one of the most common things anyone asks, we computed the curves down to week 6. This is our extension of the source data, not something the original paper endorses.
Why are there separate charts for males and females?
Because males run heavier — often 15–20% — and merging them produces a chart that's wrong for both. It's a free improvement and most puppy charts skip it.
Where does this chart come from?
We computed it ourselves from the raw dataset behind Salt et al. (2017), Growth standard charts for monitoring bodyweight in dogs of different sizes, PLoS ONE 12(9): e0182064, released CC BY 4.0. It holds 14,957,947 weight measurements from 3,475,842 dogs; 8,520,272 records survive our filters and sit behind these curves. We didn't trace the published figures — the paper only publishes its curves as pictures, with no parameter table to copy. Full derivation on our methodology page.
Should I use this chart or the calculator?
The chart if you want to look up what's typical; the calculator if you want a prediction for your puppy. The calculator follows your puppy's own percentile forward to predict its adult weight, and tells you the error range — median 15%, measured on 63,993 dogs held out of the fitting.

Related

Curves last rebuilt 2026-07-17. Computed by us from the raw dataset behind Salt C, et al. (2017), PLoS ONE 12(9): e0182064 (correction: PLOS ONE 19(11): e0314711, 2024), CC BY 4.0 — 8,520,272 weight records after filtering. Full derivation, including where we depart from the paper, on our methodology page.