Placing a puppy with a family is a decision about that one dog's life. Keeping a puppy back is a decision about the next ten years of my program. It's the highest-stakes choice I make, and I've gotten it wrong often enough to be humble about it. The "best" puppy in the litter and the right puppy to keep are not always the same dog, and learning the difference took me a long time.
This is a different decision from placing puppies with the right buyers, where the goal is matching a dog to a life. Keeping one back is about my own eye, my own program, and my own discipline not to keep too many. Let me walk through how I actually do it.
Evaluating Structure at Eight Weeks
There’s a well-known principle among breeders that a puppy at eight weeks is a fairly honest miniature of the adult it will become — before the awkward growth stages distort everything. Around seven to eight weeks, the skeletal proportions, the angles of the shoulder and rear, the length of body relative to leg, the set of the topline: these are largely laid down. After that, the puppy goes through gangly phases that hide its real structure for months.
So I evaluate structure at eight weeks, on a table, hands on each puppy, ideally with a second experienced set of eyes. I’m looking at proportion and angulation, not at the things that will change. What changes: bone and substance fill in, heads mature and broaden in the males, coats come and go, size is far from settled. What holds: the underlying architecture. A puppy that’s nicely balanced front and rear at eight weeks usually stays that way. A puppy that’s straight in the shoulder at eight weeks does not magically develop layback later.
The discipline is to trust the eight-week read and not to keep re-evaluating every week, because the adolescent phases will talk you into and out of every puppy in the litter if you let them.
Temperament Signals That Hold, and the Ones That Don’t
Structure is the easier half. Temperament is where I’ve made my worst keeper mistakes. Around seven weeks there’s a developmental window where puppies show fairly honest baseline temperament — boldness or caution, recovery from a startle, willingness to engage with a stranger, tolerance of handling and restraint. I run informal aptitude work at that age, and the baseline reactivity it reveals tends to hold.
But here’s the trap: a lot of what looks like temperament at eight weeks is actually just confidence that socialization will build anyway. The bold puppy isn’t necessarily braver for life; it may just be ahead on a curve the others will catch up on. And the genuinely valuable traits — resilience, biddability, how a dog handles pressure and recovers from it — don’t fully show until the dog is older and the socialization and early-experience work has had time to play out. The signals that hold are the extremes: the puppy that recovers instantly from everything, and the one that doesn’t recover at all. The middle is noise at eight weeks.
So I weight the baseline but I don’t worship it. A keeper has to have the structure and a temperament baseline with no red flags. Beauty does not buy back a fragile nervous system.
Why the “Pick” Puppy Isn’t Always the Keeper
The flashiest puppy in the litter — the one a show person would point to instantly — is sometimes the one I let go to a competitive home, and I keep a less obvious sibling. Why? Because what I need for my program may not be the single best individual. I might already have that puppy’s strengths in my lines and need a different one’s particular virtue — a better front, a steadier head, a specific piece I’m trying to fix. The keeper is the puppy that moves the program forward, not necessarily the one that would win on a Saturday.
This is also where honesty about my own goals matters. I stopped chasing show wins years ago, for reasons I wrote about in why I stopped showing dogs, and once that pressure was gone I got far better at keeping the useful puppy instead of the pretty one.
The Discipline of Not Keeping Too Many
The hardest skill isn’t picking the keeper. It’s keeping only one — or none. Every dog I grow on takes a slot: a slot in my house, in my time, in my future breeding decisions. Keep three from a litter “to see how they turn out” and within two years you have a kennel full of adult dogs you can’t all responsibly use, and you’ve stalled your own progress because you’re feeding the past instead of planning the future. Kennel blindness creeps in exactly this way.
My rule is to decide what I’m keeping for, decide before the litter is even born what gap I’m trying to fill, and hold myself to one. Sometimes the answer is that this litter produced no keeper and I place all of them. That’s not a failure. That’s discipline. It’s the same kind of honest accounting I keep in my breeding records, where the litters that produced no keeper are recorded as plainly as the ones that did.
The Keepers That Didn’t Pan Out
I kept a structurally gorgeous bitch puppy once, certain she was the future of my program. She grew out of it — went straight in the rear, never came back, and never produced what I’d hoped. I kept a bold, brilliant male who tested perfectly at seven weeks and developed real reactivity by eighteen months that I should have seen coming in his dam. Both taught me the same lesson: I was reading what I wanted to read. The keeper I’m proudest of, by contrast, was an unremarkable middle-of-the-litter puppy I almost let go, who simply had no faults and a temperament that never wavered. “No faults” turned out to be worth more than “spectacular.” That’s the eye you develop only by keeping the wrong ones first.
Questions People Ask Me About Choosing a Keeper
At what age can you really tell which puppy to keep?
Structure reads honestly at around eight weeks, before the adolescent growth stages distort it. Temperament baseline shows around seven weeks but only the extremes are reliable that early; the traits that matter most for a breeding program don’t fully reveal themselves until the dog is well past a year. So the eight-week pick is an educated bet, not a verdict, and good breeders stay willing to be wrong.
Should I keep two puppies if I can’t decide?
Almost never. Keeping two “to see” feels safe but it stalls your program, fills your home, and usually means neither gets the individual development a keeper needs. Make the hard single choice, place the rest in good homes, and trust that another litter will come. Keeping too many is one of the most common ways good breeders quietly lose their way.
Is the best-looking puppy the right one to keep?
Not necessarily. The right keeper is the one that advances your specific program — fills a gap, adds a virtue you lack, comes with no temperament red flags. Sometimes that’s the flashiest puppy. Often it’s a plainer sibling whose particular strength is exactly what your lines need next.