This presentation describes a QbD approach to the development and validation of robust cleaning cycles. A cleaning cycle is considered robust if it can consistently meet performance requirements under worst-case operating conditions. The worst-case operating conditions are identified through small-scale soilant characterization studies. These studies are designed to identify the operating conditions under which the system is least likely to meet the performance requirements. Thus, the QbD approach provides a higher degree of assurance that the system will meet the imposed performance requirements when it is operated within the established design space (i.e. when CPPs are within their respective operating ranges). Additional benefits are gained in streamlining process performance qualification and in optimizing continuous verification based upon process capability.