This study was designed to examine: (a) The effects of social and token reinforcement in increasing instruction-following behaviour in retarded children, where the class was addressed as a whole; (b) The effects of time-out and verbal reprimand as methods of reducing disruptive behaviour; (c) The effects of the administration of social reinforcement (verbal praise), in increasing behaviour defined as helping other children. Results show that token reinforcement is significantly more effective than social reinforcement in generating and maintaining instruction-following behaviour in a sample of severely retarded children. This supports the results of a growing body of literature relating to both the theoretical and applied aspects of reinforcement procedures. The time-out procedure was found to reduce disruptive responses at a more rapid rate than that of verbal reprimand and the effects were more lasting over time. This result was achieved using a 60 second isolation period. Verbal reprimand was found to reduce disruptive responses to slightly below the non punished level. Suggestions are made as to further developments in both procedures. The administration of verbal praise, delivered contingent upon one child helping another, led to an increase in this class of response in three children. No general conclusions can be made as to the effectiveness of this type of reinforcement in increasing helping behaviour in a group situation. The results obtained in this study further suggest that the approach taken can be successfully applied to the problem of altering the behaviour of individuals treated as a group, in a group setting. This applies to both the reinforcement and punishment conditions. The results further indicate that the systematic application of operant learning principles will both increase adaptive and decrease maladaptive behaviour in retarded children.