For my dogs I do something like this for each cue I teach: Teach the dog what sit means in a nice quiet place like a bedroom: Lure or a shape sit When I'm sure the dog will sit when prompted I add the cue while luring the sit, give the release word and the reward (if you don't give the release word before the reward the reward becomes the release to the dog). Then I do the same thing in different locations around the room Then I might work on duration in a nice quiet bedroom, lure for a sit, wait 3 seconds before releasing and reward Once the dog is consistently sitting for 3 seconds I add the cue Then I do the same thing for 5 seconds, 10 seconds, Apollo will hold a sit for 30 minuets until released. Then I do the same thing adding distractions, could be waving my arms around, could be someone entering the room, I think of ways to add distractions making the distractions more and more difficult Then I do the same thing with distance. I ask for a sit from 5 feet away, then 10 feet, then a crossed the room. Then I got to the kitchen and do the whole process over. Then the front yard, then on walks, then at Petsmart at off peak hours, then at Petsmart when it's a little busier, and so on for each of their obedience cues. I am a real stickler about my reinforcement history, or try to be, because otherwise you're in effect poisoning a cue. Unless I am SURE the dog will do it, I don't use the cue-I try to get the behavior I want another way-I hint, I lure, something. Because it's really hard to get it back on track if the dog doesn't respond to the cue, you can't go back, and if it gets too muddied and you want reliability you might as well ditch the cue that's mixed and start over with a clean cue.