The quoted "wolf %" on paper is NOT a reliable indicator of the amount of wolf in the animal, for a number of reasons.
Here is the "marbles" explanation of how "wolf percent" works. Say you have two animals: dad is a pure wolf, mom is a pure dog. On paper, their puppies are 50% wolf, 50% dog.  How it works in reality: assume "
black=wolf",  red=dog".  Put 100 black marbles in a fishbowl to represent Dad's genes: 100% wolf genes, so ALL his marbles are black. Put 100 red marbles in another fishbowl to represent Mom's genes. She is all dog, so all 100 of her marbles are red. Now, to create a puppy from their breeding: take 50 marbles at random from Dad, and 50 marbles at random from Mom. In this case, all marbles from Dad will be black, because that's all he has. All marbles from Mom will be red, for the same reason. SO, the puppy will have 50 black marbles/"wolf genes", and 50 red marbles/"dog genes". All puppies in the litter will have this combination. This is the ONLY scenario where "percent" equals "actual wolf content".

Now, take one of those puppies (let's call him Timber  :) and breed him to another puppy, from another "pure wolf X pure dog" litter just like his own. We'll call his 50% mate "Cheyenne". On paper, since Timber and Cheyenne are both 50%, their puppies will be 50 + 50 divided by 2, or, "50% wolf". However, in reality, this is what happens:
Take Timber's bowl of marbles,
50 red and 50 black. Make another bowl just like it for Cheyenne: 50 red, and 50 black. Now, they are going to create a pup. Randomly (with your eyes closed & no cheating ;) pick out 50 marbles from Timber's bowl, and 50 from Cheyenne's. Put them together in their own bowl, and look at the results. You may have randomly chosen all 50 of Timber's black marbles, and all 50 of Cheyenne's black ones as well, and come up with a pup made of all wolf genes! Not likely, but it is theoretically possible. OR, you could have chosen all the red marbles from both dogs, and produced a pure dog.  Genetically speaking, of course.
The most likely scenario is that you will get an assortment of pups, with anywhere from, say, 25 to 75 (?) "wolf" marbles. But, ALL of these pups are "50%" on paper. Obviously, the pup that is made of 75% wolf genes is going to look and act a lot more like a wolf than the one who only got 25% wolf, the "50%" label is not really accurate anymore.
Now, let's take the puppy who got only 25% wolf genes (we'll call him Lobo), and breed him to a puppy just like him from another litter. The "50%-on-paper, 25%-actual-wolf-content" female he breeds to, will be called Wolfie. ON PAPER, since both Lobo and Wolfie are 50%, their puppies are 50% as well. (50 + 50 divided by 2=50.) IN REALITY, take 50 marbles at random from Lobo's bowl, and 50 from Wolfie's. What did you come up with? It is POSSIBLE to take all 25 of each parent's black marbles, and actually get a 50% pup...but, more likely you will take, oh, 5 to 15 of those marbles from each. That makes the resultant puppy 10 to 30% wolf, not 50% like it says on his "papers". See how fast this goes downhill? There's a big difference between a 10% wolf and a 50% wolf...and this puppy is only three generations away from a  pure wolf! Most wolfdogs today are many generations away from a pure wolf, and it takes its toll on the amount of actual "wolf" in them. There are many, many "90% wolf"s, some of which are unfortunately being bred, who are almost indistinguishable from a domestic malamute or German shepherd mix.

In fact, the number of generations away from a pure wolf that a wolfdog is, will often be a better indicator of how much "wolf" he has than his "percent" will be. The "filial generation", or "F-gen/"F number" of a wolfdog simply means how many generations away from pure wolf he is. If a wolfdog is an "F-1", like Timber in our example above, it means that one parent was a pure wolf. If he is "F-2", like Lobo above, it means that a grandparent was a pure wolf. A first generation animal is probably going to be very wolflike in most respects.

If you know both the percentage AND the F-number of an animal, it gives you a better idea of its actual wolf content; however, wolf-to-dog genetics is still basically a crap shoot, and there will be wide variation amongst the pups in any given litter. Lobo in our example above is a "50% F2" wolfdog, but in this case he only inherited 25% wolf genes. One of his brothers may have inherited 75%, and therefore been a much more "wolfy" animal...yet he is still a "50% F2", just like Lobo.  The gap widens with every successive generation. This is what makes the breeding of wolf crosses such a tricky is hard enough to place a wolf cross pup with an appropriately responsible and knowledgeable owner, without being unsure of how "wolfy" the little guy will grow up to be!

There is no way to determine the amount of "wolf" in any given animal. DNA testing is underway, of course, but to date it has been unsuccessful. The difference between a wolf and a dog, genetically, is next to nothing. Phenotyping may help to give some idea of wolf content...if an animal looks exactly like
a wolf, then you are safe to assume there is some wolf blood present; however, just how much wolf is impossible to declare with any real accuracy. If the animal looks exactly like a domestic dog mix, it is safe to say that the animal is not "mostly wolf"; however, he may be *part* wolf, and there is no test to determine otherwise.

At this point in time, most knowledgeable wolfdog owners classify their animals by a scale of "wolf content" rather than by percentages. A "high content" animal is one that, for all practical purposes, looks and acts like a pure wolf. You may find slight differences in temperament and body features (bigger ears, a wider chest or heavier face). A "mid content" animal is one of obvious visible and temperamental wolf content, but who can easily be differentiated from a pure wolf. These animals range in behaviour from "mostly wolf" to "mostly dog", and many folks will describe their animal as upper-mid or lower-mid to denote which end of the "middle-wolf-content" category it falls into, genetically. "Low content" animals are, for the most part, fairly easy to pass off as a dog. They'll likely be taller, lankier, have a very intense look about them, be unusually intelligent, and tend towards a shy, expressive, alert, reactive disposition. Not always, though...low content wolfdogs vary considerably in appearance and temperament. Some act very "wolfy" despite their doggy looks, some are as tractable as domestic dogs. Most folks seem to have low content wolfdogs...which is a good thing, as on average low's make much better "pets" than upper content animals. Any breeder who tells you that a pure wolf or high-content animal makes a great "pet", and "the more wolf, the better", is one to steer clear they either do not have your best interests in mind and are only after $$, or their animals are not the content they have been represented to be. High content animals have their place, but they require a serious commitment and much effort on the part of their owner.

So, why does the wolf content "breed out" so quickly? Most 5th, 6th, 8th generation animals are barely (if at all) recognizable as wolf crosses. One would think, if the genetics are random, that there would be as many animals *over* their stated percentage, as beneath it...but it's clear that this is not the case.
Most of the reason lies with the wolfdog breeders. The majority of people out there--wolfdog fans included--are so accustomed to Hollywood's version of the wolf that they genuinely have a hard time distinguishing "wolf traits" from "husky or malamute traits". It is not uncommon for *wolfdogs* to be the actors in movies starring wolves, since wolfdogs generally lend themselves better to that sort of work than full wolves do. In addition, there is strong evidence that many of the difficult "wolfy behaviours" are linked to different facets of the "wolfy look"; and selecting the most complacent, easy-to-live-with, well-behaved, outgoing animals as breeding stock tends to diminish the wolfish looks as well. This, after all, is how we got domestic dogs in the first place! Full wolves and high contents are difficult, complicated, and downright naughty animals, and in general do not have the temperament that most folks are looking for in a "dog".
Another reason is based in what we humans often consider attractive. Many a breeder will advertise "blue eyed beauties with striking masks" or "massive parents, over 180 pounds!" None of these are "wolf" features...yet they are things that most people find appealing.
A portion of breeders will also misrepresent the wolf content in their animals, either deliberately (to justify an exorbitant asking price) or simply because they are not aware of the animal's actual content. Many will take the breeder's word that Wolfie is a "97% timberwolf"...when in reality, she is more like 97% malamute. Doesn't mean she's not a great dog! But, it's not fair to breed her to another like her, and sell the pups to people as "wolves".
There is also the possiblity of natural selection coming into play. "Use it or lose it" is a powerful force in nature, and those same forces no longer act upon wolfdog companions. They no longer need many of the features that wolves in the wild do. The same thing that allows for the possibility of genetic problems such as hip displaysia (which wolfdogs should be screened for prior to breeding, though is extremely rare in *wild* wolves), may allow for the loss of some of the "wolfy look" as well.

As always, a picture is worth a million words...go to
Wolf Park or Iowolfers for a visual description of wolf content and the effects of percent and F-gens on wolf crosses!
You can also visit
this page to see links to  some of the common dogs used in wolf/dog crosses, as well as pictures of wolves.

UPDATES! Realistic breed ID for wolf crosses can now be obtained from Embark. For a while the testing results were a poor match against actual wolf content, but the testing has come a long way.
Also, I miss the old Wolf Park pages on Identification of a Wolfdog, the poster presentation...they were so good! So, I've re-hosted the first two here. Wolfdog ID (page 1) and Percentages (page 2)

~Good puppy, Go HOME!~