Recently, there has been increased attention toward determination of the grain boundary character distribution (GBCD) and manipulation of the relative fractions of "special" and "random" boundaries in the microstructure through thermomechanical processing techniques in order to improve materials properties like corrosion, creep resistance, and weldability. Most of the "optimization" treatments reported in the literature have been performed on f.c.c. metals and alloys with medium to low stacking fault energies, and have resulted in microstructures with high fractions of Sigma 3, Sigma 9, and Sigma 27 boundaries, or Sigma 3(n) boundaries. Experiments to modify the GBCD of oxygen-free electronic Cu and Inconel 600 through sequential thermomechanical processing are presented and the efficacy of these processing routes is assessed in terms of microstructural descriptors like the random grain boundary network and the distribution of triple junctions. This analysis has shown that Sigma 3(n) reactions, which occur as a consequence of multiple twinning, not only improve the GBCD but are also critical for disrupting the connectivity of the random boundary network. Published by Elsevier Science Ltd on behalf of Acta Metallurgica Inc.