Previous: Part 1: Understanding Your Shiba
This is Part 2 of 6. Part 1 explained why generic training fails on Shibas. This part covers the actual commands: name recognition, sit, down, stay, come, leave it, loose-leash walking, and go to your spot. Every progression is adapted for the Shiba brain.
Before you start
Three things need to be true before you run a training session.
The dog's health is good. A Shiba in pain will not train. If your dog has been off lately (refusing food, slow to stand, flinching when touched, scratching at ears), see the vet before you blame the dog for failing a stay. Ear infections, hip pain, dental issues, and thyroid problems all surface as training resistance first. The vet visit costs less than two months of frustrating sessions.
The environment is set up. Train in a low-distraction space first. That means your living room with the TV off, no other pets in the room, door closed. A Shiba cannot learn a new behavior while a squirrel is sitting on the back fence. The dog's entire brain will be tracking that squirrel, and nothing you say will register. Set the dog up for success by removing the competition. Outdoor training comes later.
Your treat hierarchy is built for THIS dog. Part 1 walked through the three levels: Level 1 (kibble, for easy behaviors), Level 2 (training treats, cheese, hot dog pieces, for new behaviors), Level 3 (freeze-dried liver, chicken breast, whatever your dog loses their mind over, reserved for the hardest asks). Bring all three to every session. If you are using Level 3 for everything, the value deflates when you need it most.
And as Part 1 covered: sessions are 5 to 10 minutes. End before the dog checks out, not after.
Name recognition
The most underrated skill in dog training. If your dog does not look at you when you say their name, nothing else works. Recall fails. Leave it fails. Stay fails. Everything downstream of attention is a coin flip.
Shibas hear their name and weigh whether responding is worth the effort. Your job is to make name response a reflex.
Step 1: Name equals treat. Hold a Level 2 treat. Stand near the dog. Say their name once, neutral tone. The moment the dog turns toward you, mark with "yes" and deliver. Repeat 5 times, end the session.
Step 2: Catch them not looking. Wait until the dog is sniffing the floor or looking out the window. Say their name once. When they turn, "yes" and treat. If the dog ignores you, do not repeat the name. Walk somewhere else and try again later. Saying the name twice teaches the dog that the first one is optional.
Step 3: Build duration. Once the turn is reliable, start rewarding for holding eye contact. Say the name, the dog looks, wait one second before marking. Build to two seconds, then three.
Step 4: Add distraction. Practice in different rooms, the backyard, on walks. Each new environment is a fresh skill. Drop difficulty (higher-value treat, shorter distance) every time you change context.
The rule: never say the dog's name as a correction. "Kenji! Stop that!" teaches the dog their name means trouble. Save the name for one message: "I have something good for you."
Sit
The easiest basic command for Shibas. No vulnerability required, fast payoff.
Step 1: Lure. Hold a treat at the dog's nose, move it up and back over the head. The rear goes down as the head tracks up. The moment the rear hits ground, "yes" and deliver. Repeat 5 times. If your Shiba does not follow the lure (some back up instead), skip to step 2.
Step 2: Capture. Stand near the dog with treats in your pocket. Wait. When the dog sits on their own, mark "yes" and treat immediately. Do this for two days. The dog starts offering sits to make treats appear. Capturing is slower but produces a more reliable behavior because the dog thinks it was their idea.
Step 3: Name it. Once the dog offers sits reliably, add the cue. As the dog is about to sit, say "sit" once. Mark and treat when the rear hits ground. After 10 to 20 reps, the word sticks.
Common Shiba issue: the dog hovers without fully sitting when the treat appears. Withhold the treat until the rear is firmly on the ground.
Down
The hardest basic command for Shibas. Lying down exposes the belly and signals vulnerability. A Shiba who does not trust the environment will not lie down, no matter how good the treat. That is survival instinct, not stubbornness.
Never push the dog down physically. Do not push shoulders, step on the leash, or manipulate the body. Those methods destroy the Relationship Checking Account (Part 1) and produce a dog who hates the down position permanently.
Step 1: Lure from sit. Hold a Level 3 treat at the dog's nose, slowly lower it straight down between the front paws. If the dog follows all the way down, "yes" and deliver. If the dog only drops partway, reward the partial.
Step 2: Shape it. If luring does not work, reward incremental approximations. The dog dips their head: treat. Leans forward: treat. Bends one elbow: treat. Elbows touch the ground: jackpot (three treats rapid-fire with praise). Whole body drops: jackpot and party. Shaping takes patience. It also works on the Shibas that luring does not reach.
Step 3: Name it. Add the cue "down" as the dog is going into position. Mark once the belly hits the ground.
Step 4: Build duration. Reward at 2 seconds, then 5, then 10, then 30. If the dog pops up early, cut the time in half.
Common pattern: the dog downs at home but refuses in public. Different surfaces and spaces feel less safe. Rebuild from scratch in each new context.
Stay
The most useful everyday skill. A solid stay keeps the dog safe at the front door, calm at the vet, and out of the way during groceries. The framework is the "three D's": duration, distance, distraction. Train in that order, one at a time.
Step 1: Duration. Ask for sit or down. Wait 1 second. Mark and treat. Build to 2, 5, 10, then 30 seconds. Stay within arm's reach the whole time.
Step 2: Release cue. Pick a word ("okay," "free," "break") that means "you can move." Use it every time. Without a release cue, the dog decides when stay ends.
Step 3: Distance. After the dog holds 30 seconds with you right there, take one step back. Return. Mark and treat. Build to two steps, five, across the room. Always return to the dog to release. Do not call the dog out of a stay, or stay and recall become the same command and both break.
Step 4: Distraction. Drop a treat near the dog, have someone walk through the room, open the front door. Each distraction is a fresh challenge. Lower duration and distance when you add a new distraction, then rebuild.
Common Shiba pattern: the dog holds the stay when you face them and breaks when you turn your back. Fix by practicing back-turning as a specific distraction in step 4.
Come (recall)
THE Shiba-specific challenge. Recall is the most important skill you will teach and the hardest to make reliable. When a Shiba locks onto prey, your voice becomes background noise. The chase instinct is older and louder than anything you have trained.
Perfect off-leash recall on a Shiba is rare. A reliable long-line recall that keeps the dog alive is achievable in 6 to 12 months.
(a) Puppy following instinct indoors. Crouch a few feet from the dog, open your arms, call their name plus "come!" in a warm tone. When the dog arrives, party: treats, praise, play, the whole show. Make it the best 10 seconds of the dog's day. Repeat 5 times per session.
(b) Name + come in a boring room. Move to a longer hallway. Call from another room. Same party every time. If the dog does not come, walk to them, gently lead them to your spot, then party. Never punish a slow recall.
(c) Come away from mild distraction. The dog is sniffing a toy. Call "come." When the dog leaves the distraction, Level 2 treat and praise. If the distraction wins, it is too strong for this stage.
(d) Long-line recall in the yard. Attach a 15- to 30-foot line to the dog's harness (not collar; collars can hurt the neck on a fast recall). Let them sniff and explore at the end of the line. Call. If they come, Level 3 treats and a full party. If they do not come, reel the line in gently (no yanking) so the dog arrives at your feet, then reward anyway. Then release the dog back to sniffing as a second reward. This is the Premack Principle in action: the dog does the less-desirable thing (come back to you) to access the more-desirable thing (freedom to go sniff again). Come does not end the fun. Come is the gate that opens more fun. This reframe is what makes recall work on a Shiba long-term.
(e) Recall with a helper holding the dog. A friend holds the dog by the harness. Walk 20 feet away. Call. The friend releases when the dog pulls toward you. The dog sprints to you. Party. Builds drive toward the recall word.
(f) Recall in a fenced area with real distraction. Off-line in a fully fenced space. Let the dog explore. Wait until they are interested in something. Call. If they come, Level 3 treats and celebration. If they do not, clip the line back on and return to step (d) for a few more weeks.
The cardinal rule: NEVER call your dog for something unpleasant. Do not call them for baths, nail trims, crating, or leaving the park. Walk to the dog for those things. Save the recall word for moments where the dog gets a guaranteed party or gets released back to what they were doing. One bad outcome after a recall can poison weeks of work. The dog remembers. "The last three times I came when called, something unpleasant happened. Why would I come again?" The Relationship Checking Account applies directly: every recall that ends well is a deposit, every recall that ends badly is a withdrawal. Make the math overwhelmingly positive.
Leave it
Critical for prey drive management. The same instinct that overrides recall sends a Shiba lunging for chicken bones, dropped medication, and dead rodents. A reliable "leave it" is emergency equipment, not a trick. It is a lifetime contract.
Step 1: Closed fist. Hold a Level 1 treat in a closed fist. The dog will sniff, lick, paw. Wait. The instant the dog backs off, "yes" and deliver a Level 2 treat from your other hand. The closed-fist treat is never the reward. Repeat 10 times.
Step 2: Open hand. Palm open, treat visible. If the dog lunges, close the fist. When the dog backs off from an open palm, "yes" and reward from the other hand.
Step 3: Treat on the floor, covered. Place a treat on the floor, cover with your hand. Same drill. Dog backs off, gets a better one from your pocket.
Step 4: Add the cue. Say "leave it" right before presenting the treat. After 15 to 20 reps, the verbal cue is established.
Step 5: Treat uncovered on the floor. Place the treat, say "leave it," stand ready to cover with your foot. When the dog backs off without you covering, mark and reward from your hand. The floor treat stays on the floor.
Step 6: Drop while walking. Walk past the dog, drop a treat. "Leave it." If the dog ignores the floor treat, mark and reward from your hand. If the dog goes for it, step on it before they grab it. No correction, just block and reset.
Step 7: Real-world items. Practice with a sock, bread on the sidewalk, a dead bird on the walking path. Generalize so "leave it" means "do not touch that thing, whatever it is," not just "ignore the training treat."
Loose-leash walking
The most common complaint we hear. Every time the dog reaches the sniff spot by pulling, the pull is reinforced. By the time most owners contact us, the dog has been rewarded for pulling thousands of times.
Non-negotiables:
No retractable leashes. They teach the dog that pulling extends the leash. Use a 6-foot flat leash.
No prong collars, choke chains, or e-collars. The research is clear: increased fear, increased aggression, suppressed warning signals before bites.
Use a martingale collar. Shibas have thick necks and narrow heads and can slip standard flat collars with one backward pull. A martingale tightens slightly when the dog pulls (not enough to choke, just enough to prevent escape) and loosens when the leash is slack. It is the breed-standard collar recommendation. Pair it with a front-clip harness.
Front-clip harness. Leash attaches at the chest. When the dog pulls, the harness redirects them sideways, breaking forward momentum. Not painful, not punishing, just mechanically effective.
The method: stop when they pull. Walk forward. The moment the leash goes tight, stop. Stand still. Do not yank. When the dog releases tension (steps back, turns to look at you), "yes" and walk again. Pulling stops the walk, slack restarts it.
Variant: instead of stopping, turn 180 degrees. The dog must follow. After enough reps, the dog starts watching you because the direction is unpredictable.
The first week may feel like 50 starts and stops per walk. That is the process working. Within a month of genuine consistency, most Shibas walk with reasonable slack.
Go to your spot
The most useful "stay out of the way" skill. Go to a specific location, lie down, stay until released. The biggest quality-of-life improvement most owners report in their first month. The spot can be a bed, mat, rug, or crate. If using a crate: the crate is a sanctuary, not punishment. Never send the dog to the crate as a consequence.
Step 1: Generate interest. Place the bed on the floor. Toss a treat onto it every time the dog interacts with it (sniffs, steps near, looks at it). Five minutes, end the session.
Step 2: Reward stepping on. Only mark when the dog puts a paw on the spot. Build to four paws.
Step 3: Reward lying down on it. Withhold the treat until the dog lies down on the spot. Most dogs figure this out quickly because down is already trained.
Step 4: Add the cue. As the dog heads to the spot, say "place" (or your word). Mark and treat once in position.
Step 5: Send from 2 feet. Cue from a short distance. Build to 5 feet.
Step 6: Send from across the room. The dog goes to the spot from wherever you send them.
Step 7: Build duration. Start at 5 seconds on the spot, build to 1 minute, 5 minutes, 15 minutes. Reward intermittently while the dog holds position.
Step 8: Add distraction. Practice while the doorbell rings, guests arrive, dinner is cooking.
Step 9: Real-life integration. Cue the spot when the doorbell actually rings, when groceries come in, when company arrives. Within 2 to 3 months, the dog starts heading to the spot on their own when they hear the triggers. That is the win.
Session structure
Rules that apply to every command above.
5 to 10 minutes, maximum. Covered in Part 1: past 10 minutes, the Shiba's tolerance for repetition is exhausted. Sessions can be twice daily, but each one is short.
One command focus per session. Do not try to chain sit, down, and stay together while the dog is still learning each one. Build each behavior to fluency before stacking. Once each is solid on its own, mix them in real-life practice.
End on success. The last rep of every session should be one the dog gets right. If a session goes badly (the dog is distracted, you are frustrated, the timing is off), drop the difficulty and end on an easy win. Ask for something the dog already knows, reward it, and stop. The dog's last memory of the session shapes their attitude toward the next one.
Integrate into daily routines. The most effective training happens outside of formal sessions. Sit before every meal. Stay while getting the mail. Recall before the leash goes on for a walk (so the dog associates coming to you with going somewhere good, not with losing freedom). Leave it when passing food on the counter. Place when you are sorting mail or on a phone call. The dog learns that the commands are part of daily life, not a special exercise that only happens in that 5-minute training block. A Shiba who only performs in training mode is a Shiba who has not generalized the skills. Real life is where the generalization happens.
Track progress. A note on your phone works. "Day 1: sit reliable indoors, no distraction. Day 5: sit reliable in the yard. Day 14: sit reliable at the park with a dog 50 feet away." You will lose track of how far you have come without a written record. On the bad days (there will be bad days), the log reminds you that progress is real.
Foundation skills are not the whole picture. Part 3 covers socialization, including the specific challenge of socializing an adult rescue Shiba who missed the critical puppy window. If you have a rescue who reacts to strangers or other dogs, that is where to go next.
This is Part 2 of the NorCal Shiba Training Guide. Part 3: Socializing a Shiba covers socialization for puppies and adult rescues, including the decompression protocol and counter-conditioning. Have questions? Email norcalshiba@gmail.com or text (707) 718-3579. If your Shiba needs hands-on training, our training programs are built for this breed.