One of the most important aspects of treatment outcome research is establishing treatment integrity. Integrity of the treatment refers to the degree to which treatment is implemented as intended. Research examining the relationship between treatment integrity and therapeutic change has produced conflicting results. However, assessment, design strategies, and the possible confound of integrity with other variables may explain the inconsistency in findings. This paper elaborates the limitations of existing strategies for evaluating the relationship between treatment integrity and outcome. Recommendations for future research include controlling possible confounding variables, experimentally manipulating treatment integrity, and using novel assessment and evaluation strategies.