In accordance with the present invention, a training device is provided for training animals to perform or refrain from performing a target activity. The training device includes a reward dispenser that is capable of dispensing a reward to an animal being trained upon receiving a reward dispensing command. The reward dispensing command includes a hopper for holding a quantity of rewards. The hopper includes an entrance port for receiving rewards and an exit port through which rewards can be dispensed. The reward dispenser also includes a delivery assembly for controllably delivering rewards from the hopper to the exit port, the delivery members being placeable in contact with the rewards to be delivered. A motor member is provided that is capable of maintaining the delivery members in a static position to prevent the delivery of a reward, and moving the delivery member to deliver at least one reward to the exit port. The delivery member includes at least one bristle-containing member. A controller is provided that is capable of receiving a command and actuating the motor upon receipt of the command to move the delivery member to thereby dispense a reward to the exit port.