GrowthGuide

Double at 16 weeks? Puppy weight rules tested on real data

The question

Does the ‘double the weight’ rule work for my size of dog?

Short answer

Only for small dogs, and only on average. Doubling the weight at 16 weeks lands on the median for toy and small puppies but runs 10% to 29% low for medium to giant dogs, and doubling at 6 months overshoots every size by 29% to 46%. Even the best rule put fewer than half of puppies within 10% of their adult weight.

Updated  · sources read  · Data and sources →

Ask around about how big a puppy will get and someone will offer a rule: double the weight at 16 weeks, double it at 4 months, double it at 6 months. They are easy to remember, and they sound precise. This guide checks six of them against real puppies, size by size. None fits every size, one fits none, and even the best of them puts fewer than half of puppies within 10% of their adult weight.

The test data are the open, CC BY 4.0 visit records behind a 2017 study of more than 6 million young dogs seen at U.S. veterinary hospitals. Everything was read and computed on October 7, 2026, with the same code our calculators use.

How the rules were checked

For each puppy in the data with a weight recorded within a week of the rule’s age, we applied the rule — for example, its weight at 16 weeks times 2 — and compared the answer with the same dog’s actual adult weight, taken as its median weight at 2 to 3 years old. We used only weights from healthy or preventive-care visits, from dogs of the 31 breeds in the data that were never recorded as thin or heavy and had a normal body condition at 2 to 3 years.

That gives every dog an error: +10% means the rule guessed 10% too heavy, −10% means 10% too light. We then summarized the errors for each of the six size classes the study used, by expected adult weight: class I under 6.5 kg (14.3 lb), II 6.5–9 kg, III 9–15 kg, IV 15–30 kg, V 30–40 kg and VI 40 kg (88.2 lb) and over. Each cell rests on between 524 and 4,276 dogs.

The results, size by size

The first table shows the median error: the miss for a typical puppy. A rule that is right on average scores 0%.

RuleI (toy)II (small)III (medium)IV (medium-large)V (large)VI (giant)
2 × the weight at 16 weeks0%−1%−10%−17%−19%−29%
2 × the weight at 4 months (17.4 weeks)+8%+8%−1%−9%−9%−20%
2 × the weight at 6 months+44%+46%+40%+38%+38%+29%
2.5 × the weight at 14 weeks+11%+7%−4%−14%−15%−29%
3 × the weight at 12 weeks+15%+9%−4%−16%−17%−31%
4 × the weight at 8 weeks+11%−2%−16%−31%−35%−49%

A median of 0% does not mean the rule works for your puppy. The second table shows how many dogs each rule placed within 10% of their real adult weight.

RuleI (toy)II (small)III (medium)IV (medium-large)V (large)VI (giant)
2 × the weight at 16 weeks35%42%37%26%24%6%
2 × the weight at 4 months (17.4 weeks)33%38%42%39%42%21%
2 × the weight at 6 months7%4%7%7%7%13%
2.5 × the weight at 14 weeks28%34%37%30%30%11%
3 × the weight at 12 weeks26%31%34%28%28%8%
4 × the weight at 8 weeks24%30%26%12%8%2%

What the tables say:

  • Doubling at 16 weeks is centered for toy and small dogs, but the spread is wide: for class I, 8 in 10 puppies had errors between −26% and +30%. For large dogs it runs low almost every time — in class V, 8 in 10 errors fell between −37% and −1%.
  • Doubling at 4 months is the least-bad rule for medium to large dogs, yet still lands within 10% for only about 4 in 10 of them.
  • Doubling at 6 months is wrong for every size. A typical puppy at 6 months is already 64% to 73% of its adult weight, not 50%, so the rule overshoots by 29% to 46%, and only 4% to 13% of dogs land within 10%.
  • The early rules (×3 at 12 weeks, ×4 at 8 weeks) drift from too heavy for toys to far too light for giants.

The straight-line rule many calculators use

Some calculators skip the folk rules and divide today’s weight by the age in weeks, then multiply by 52 — as if a puppy gained the same amount every week until its first birthday. Puppies don’t grow that way: the Merck Veterinary Manual describes rapid growth for the first 5 months and a plateau beginning after 6 months.

For the typical puppy in each class, here is how far the straight-line answer overshoots:

Size class12 weeks16 weeks20 weeks6 months9 months
I (toy)+67%+63%+57%+43%+11%
II (small)+58%+61%+57%+46%+16%
III (medium)+39%+47%+47%+40%+13%
IV (medium-large)+22%+35%+39%+37%+15%
V (large)+20%+32%+37%+37%+15%
VI (giant)0%+15%+25%+28%+13%

It is worst for toy and small puppies in their first months, which is exactly when owners of small dogs most want an answer. A 1.5-lb (680-g) Chihuahua at 12 weeks comes out at 6.5 lb by the straight line; the data-based estimate is 3.9 lb, with a middle half of 3.3 to 4.7 lb.

Why no single rule can fit

Every rule is a hidden claim about one number: the share of adult weight a puppy has at a given age. Doubling at 16 weeks assumes 50%. In the U.S. data, the typical share at 16 weeks runs from 50% in toy and small dogs down to 35% in giants. A rule that is right for a Chihuahua has to be wrong for a Great Dane.

The 2004 kennel study by Hawthorne and colleagues, which weighed puppies of 12 breeds every week, shows the same pattern. Working its published breed curves forward to 16 weeks gives anywhere from about 32% for English Mastiffs to 68% for Papillons. Its authors noted that the period of fast early growth ended at about 11 weeks in toy breeds but lasted until about 5 months in giant breeds.

And even within one size class, puppies differ. Some are early growers and some are late ones, so any one-number rule will miss a large share of individual dogs, whatever the multiplier.

What to use instead

Use the share for your puppy’s size class and age, and keep the range.

Take a male Labrador weighing 25 lb (11.3 kg) at 16 weeks. Doubling says 50 lb. The data say a class V puppy at 16 weeks has typically reached 41% of adult weight, which gives about 61 lb, with a middle half of 55 to 69 lb and 8 in 10 between 51 and 79 lb. The doubling answer falls below even that wider band.

Now a male Golden Retriever weighing 45 lb at 26 weeks. Doubling at 6 months says 90 lb. The data give about 66 lb, with a middle half of 60 to 73 lb and 8 in 10 between 56 and 81 lb. This time the rule lands above the band.

Weigh again every few weeks: a run of estimates that agree tells you far more than any single rule. The method, with a worked example and the full table, is in how big will my puppy get. If your puppy’s weight stops rising, or jumps well above or below its earlier path, talk to your vet.

Bottom line

The doubling rules are not useless, but each fits only part of the size range and none is reliable for an individual puppy. Doubling at 16 weeks suits toy and small dogs on average; doubling at 4 months suits medium to large dogs on average; doubling at 6 months suits no size. For a number with an honest range, enter your puppy’s weight and age in the puppy weight calculator, then track it on the puppy growth chart.

Calculators for this guide

Sources

Each source was read on the date shown. When one changes, this guide and its date change with it.

  1. Growth standard charts for monitoring bodyweight in dogs of different sizesPLOS ONE (Salt C, Morris PJ, German AJ, Wilson D, Lund EM, Cole TJ, Butterwick RF) · 2017;12(9):e0182064, doi:10.1371/journal.pone.0182064 · CC BY 4.0 · read
  2. Growth standard charts for monitoring bodyweight in dogs of different sizes - SUPPORTING DATAUniversity of Liverpool DataCat 377 · Final_Data_PLOS.zip, metadata revision 7, doi:10.17638/datacat.liverpool.ac.uk/377 · CC BY 4.0 · read
  3. Body-weight changes during growth in puppies of different breedsThe Journal of Nutrition (Hawthorne AJ, Booles D, Nugent PA, Gettinby G, Wilkinson J) · 2004;134(8 Suppl):2027S–2030S, doi:10.1093/jn/134.8.2027S (full text read in a real browser) · read
  4. Feeding Practices in Small AnimalsMerck Veterinary Manual · Full review Nov 2023; modified Oct 2024 · read

Data licences

Growth bands derived by Puppy Scale from Salt C et al. (2017), PLOS ONE 12(9):e0182064, data: University of Liverpool DataCat 377, licensed CC BY 4.0. Our tables are empirical summaries, not the published model curves. Changes: see the method.

Guides that pick up where this one stops.