Introduction: Eating disorders are complex pathologies that show a variable success in specialized treatment processes. For this reason it is relevant to count with useful and standardized instruments in order to measure the evolution of these patients. Methodology: Quantitative, descriptive and relational design. Clinical files of 26 patients were reviewed and the EAT-26 was applied in two instances to 13 patients diagnosed with ED by a specialized treatment unit. Results: The total scale EAT-26 so as the Control dimension vary in a significant way during treatment. The Bulimia subscale changes as well, and although the change is not significant, it is progressive to the extent that treatment unfolds. Conclusion: The EAT-26 could be a useful instrument to measure change in patients diagnosed with ED and treated in specialized units.