Counted from the first mating, the 290 pregnancies in Concannon and colleagues' 1983 Beagle colony ended in whelping (giving birth) on 16 different days, from day 57 to day 72. Counted from the peak of luteinising hormone (LH), the surge that sets off ovulation, the 54 pregnancies timed that way in the same colony ended on just 3, days 64 to 66. The abstract prints the averages as 65.3 ± 0.2 and 65.1 ± 0.1 days without saying what the ± measures. We infer they are standard errors, since a standard deviation that small cannot fit those ranges, and on that reading the spread is about 3.4 days from mating against under 1 from the LH peak (our arithmetic). The familiar "63 days" is the American Kennel Club's average counted from ovulation.
Ovulation anchors the familiar 63-day figure
The AKC's page on the reproductive cycle, updated 18 December 2025, gives "63 days on average from ovulation". Its figures from the LH surge, the progesterone rise and the first breeding all differ, because the starting events differ.
Concannon's 2002 lecture to the World Small Animal Veterinary Association (WSAVA) puts ovulation "about 2-2.5 days after the preovulatory LH surge". The eggs then spend another two or three days maturing in the oviducts before they can be fertilised. Counted from conception, a pregnancy is therefore shorter than 63 days. No source we opened gives that figure, so we give none.
Week-long sperm survival cuts the mating date loose
Single matings three days before the LH peak produced pregnancies in the 1983 colony, which the authors read as a sperm fertile life of "at least 6 days". Concannon's 2002 lecture gives up to 8 days, and his 2004 lecture with Verstegen up to 9. In the 1983 abstract, every interval of 61 days or less followed a mating at least 3 days after the LH peak, and every one of 68 days or more followed a first or single mating at least 2 days before it. Even day 72 is no outer limit: the 2004 lecture gives "72 or more" from the first of several matings and "as short as 55 or 56 days" after a late single one.
Until 14 October 2026 our dog pregnancy due date calculator asked for the first of several matings, called day 63 the due date and closed its window at day 68, while the Method and sources block on the same page said 57 to 72. On a normal approximation of the colony's first-mating spread (our arithmetic), that cut-off would have left between one whelping in ten and one in four after the window. The version revised that day asks which event the date marks. From a first mating it gives days 57 to 72. From an LH peak it gives 64 to 66, Concannon's colony figure; counted from a progesterone rise, the same three days held only two thirds of the dogs in Kutzler's clinic study below, so its window from a progesterone rise runs from 62 to 68. Its ovulation window of 62 to 64 days is labelled as a subtraction, and the middle date as the middle of the range.
A progesterone-timed mating pins the average near day 61
Okkens and colleagues (2001) mated 113 female dogs of six breeds once each, on a day chosen by measuring progesterone in the blood. Gestation ran from 58 to 65 days, with a mean of 61.4 and a standard deviation of 1.5. Concannon's colony, counted from the first mating, averaged 65.3, or 3.9 days later.
Most of that gap is timing. Against the Beagles' 65 days from the LH peak, Okkens' single mating sits about 3.6 days after the peak, on our arithmetic, which is where ovulation and egg maturation put the fertile days. Some of it is breed. Eilts and colleagues (2005) counted 764 litters from 308 female dogs from the initial progesterone rise and found Labrador retrievers averaging 62.9 days and German shepherds 63.6, both below the Beagles. Okkens' own Labradors averaged 60.9 days from the mating and its German shepherds 60.4. Compared breed for breed, the timed mating fell about 2.0 days after the progesterone rise for Labradors and 3.2 for German shepherds. That is our arithmetic across two laboratories and two slightly different day zeros.
The hormone surge ties whelping to a three-day band
All 54 of the Beagle pregnancies timed from the LH peak in 1983 ended on day 64, 65 or 66 counted from that peak. Concannon's 2002 lecture calls pregnancy length in dogs "remarkably consistent" on this count, "in nearly all instances". That is data from a research colony.
Kutzler and colleagues (2003) tested the same idea retrospectively on clinical records: 63 female dogs of 19 breeds, with whelping predicted at 65 days after the initial progesterone rise, an event whose timing, they write, "cannot be differentiated" from the LH peak. The prediction fell within a day of the actual date for 67% of them and within two days for 90%. Within three days it caught all 63. The three-day band holds in one closely watched colony, and in a clinic it took a seven-day band to catch every whelping.
Breed and litter size loosen the mooring by a day or more
Okkens' breed averages ran from 60.4 days for German shepherds to 62.8 for West Highland white terriers. The abstract gives means for only four of its six breeds, so 2.4 days is the least the breed averages differ by. The authors conclude, "although it was not conclusively proven", that breed is "a major determinant" of gestation length, perhaps through litter size. Within any one breed their dogs still varied by 4 to 7 days.
Eilts found that litters of four or fewer puppies ran longer, by about a day on average, and that a female dog's age and number of earlier litters made no difference. Romagnoli's 2006 WSAVA lecture notes that these hormone timings were "studied mostly in Beagles" and that data on other breeds are lacking. None of the studies we opened measured small companion breeds such as the Shih Tzu, the Poodle or the Maltese.
Each reference event tethers its own due-date window
Romagnoli's 2006 lecture sets four day zeros side by side: a single breeding, the LH surge, ovulation and the onset of cytological dioestrus, a stage read from cells in a vaginal smear. The version below adds the progesterone rise and the timed single mating.
| Day zero | Days to whelping | Source and scope |
|---|---|---|
| First mating | 57–72, mean 65.3 | Concannon et al. 1983, 290 Beagle pregnancies; 55 or 56 after a late single mating and 72 or more from the first of several (Concannon and Verstegen 2004) |
| Single mating timed by progesterone | 58–65, mean 61.4 | Okkens et al. 2001, 113 female dogs of six breeds |
| LH peak | 64–66, mean 65.1 | Concannon et al. 1983, 54 Beagle pregnancies; also Romagnoli 2006 |
| Initial progesterone rise | 65 predicted; within 1 day for 67%, within 3 for all | Kutzler et al. 2003, 63 female dogs of 19 breeds; the AKC gives 64–66 |
| Ovulation | 62–64, about 63 on average | Romagnoli 2006; average from the AKC |
| Onset of cytological dioestrus | 56–58 | Romagnoli 2006 |
View author profile → · Editorial policy
