Abstract The increasing use of token economies has been warranted by its impressive results in altering a wide range of behaviors with diverse treatment populations. However, a close examination of the literature reveals a relatively consistent percentage of treatment failures. Evidence is reviewed which demonstrates that the failure of some patients to respond to the token economy has laboratory precedents. Possible explanations for unresponsiveness of patients to token reinforcement are entertained. The implications for the operant paradigm represented by these failures are discussed. Finally, several solutions are suggested to decrease the number of patients who fail to respond.