Although the principles governing the application of positive reinforcement to classroom situations remain clear and theoretically consistent, practical problems arise during application that are not easily anticipated. This paper focuses on one practical issue: the selection of appropriate reinforcers to use in inner city high schools. The paper is based on experiences in the development of reinforcement programs in a number of high schools in an urban area. Grade related reinforcers were found to be effective in modifying student behaviors. Nontangible reinforcers were found to be as effective as tangible reinforcers. Token systems were effective; the tokens became reinforcing in and of themselves. Spontaneous social reinforcement could be the most effective system, if it is use properly. (Author/MK)