In this work, an experimental design methodology was applied to optimize the degradation of an Orange II (OII) solution, a non-biodegradable azo dye, while minimizing also the leaching of iron from the catalyst support in a heterogeneous Fenton-like process. The independent variables considered were the temperature, H2O2 concentration, and catalyst (iron-impregnated pillared saponite clay) load. The multivariate experimental design allowed the development of empiric quadratic models for dye degradation, TOC removal, and iron leaching after 1, 2, 3, and 4 h of reaction, which were adequate to predict responses in all of the range of experimental conditions used. Data obtained revealed that the heterogeneous Fenton-like process is promising for the degradation of the studied azo dye. Actually, after 4 h oxidation color removals near 100% and TOC reductions of at least 65% were experimentally achieved when the temperature was 40 degrees C or higher. Iron leaching was also quite small after 4 h of oxidation (in the range 0.66-5%), pointing to a good stability of the catalyst. Besides, the optimal conditions depend on the response factor considered, being advisable to use less-aggressive conditions if responses are taken at longer reaction times. Particularly, temperature, but also catalyst concentration, were found out to be the main parameters affecting all of the responses (dye degradation, TOC removal, and iron leaching), whereas the effect of the initial H2O2 concentration was found to be negligible. Finally, the process was optimized considering the three responses simultaneously, allowing defining optimal regions for the significant process variables (temperature and catalyst dose in the slurry batch reactor).