Abstract: A learning device for attention deficit children is disclosed, having a wall mounted cabinet with an audio equipped model train located in a mirrored frontal surface, and sufficient timing devices to guide the child through a series of planned tasks, and having storage capability for associated task cards, listing pre-planned tasks, to be completed by the child, and storage space for negative correction cards, to be given to child upon non-completion of tasks. A slot is located on the side of the cabinet, and, when child successfully completes a task, in the time allowed, insertion of the corresponding card in the slot ejects a chip to the child, to be accumulated by the child, and, at certain periods of time, marked on a furnished calendar, the child is rewarded from a list of chosen rewards, each reward requiring a different number of chips, the rewards being weighted by importance to child.