This Monograph reports the results of teaching preschool teachers to be successful at increasing desired behaviors of their children, thus becoming successful teachers. Five teacher-training techniques were examined experimentally under single-subject designs: written assignments, feedback from viewing graphs, on-the-spot feedback from a wireless radio (Bug-in-the-Ear), feedback from an observer, and self-counting. Those teaching procedures that included prompt and frequent information to the teacher about the behavior under study were the most effective techniques. Self-counting, in which the teacher tallied the number of times she emitted the behavior of either priming or reinforcing social or verbal behavior of a child (or children), and observer feedback, in which the observer reported to the teacher periodically during the hour the frequency of her behavior, were the most reliable teaching techniques. The other procedures, while less reliable than self-counting and observer feedback, were effective with some teachers. Maintenance of teacher behavior across settings was examined with a group of Head Start teachers, and maintenance of teacher behaviors across different child behaviors and different children was examined with three student teachers. The results indicated that teaching was more likely to maintain if it occurred in the teacher's home setting rather than at another site. In all cases, when generalization occurred across settings, time, or children, the frequency of the teacher's behavior was not as high as when the relevant behavior had been trained directly. Results supported the proposal that it is possible to define effective teacher behavior, not just characterize it, as it occurs in the classroom, and that effectiveness can be measured by defining and observing the child behaviors to which teacher behaviors are directed.