This paper summarizes the development of an optimized Cu(II)-catalyzed oxidation process for the dimer formation of ApoA-1M, the protein component of an HDL mimic with potential efficacy in atherosclerotic plaque reduction. This oxidation represents a major processing step in the large-scale manufacture of ApoA-1M dimer. The development approach utilizes reaction kinetic modeling, design of experiment, and neural network analysis to establish a quantitative basis for specification of process operating ranges. This work is thus an example of quality by design (QbD) principles where a systematic mechanistic understanding combined with multivariate analysis is integral to process development. Operating within the defined optimum window of operation resulted in both enhanced performance and robustness for the oxidation step in the piloting and manufacturing campaigns. (C) 2009 Elsevier Ltd. All rights reserved.