High-fidelity three-dimensional (3-D) simulations of multiphase flow and mixing in dilute bubble swarms were performed using the Euler-Lagrange simulation approach. Included was species transport, as well as complex chemical reactions in the simulations. It was found that the algebraic SGS model satisfactory predicts experimental data for the mean flow field. A detailed description of multiphase flow was used and developed to simulate the time evolution of scalar and reactive mixing in a bubble column. An analysis involving the scale of segregation Phi, a metric that characterizes the mean driving force for mixing, is applied for the first time to multiphase flow. The study shows that Phi is inversely proportional to the bubble diameter at constant gas-feed rate, but only a weak function of the gas-feed rate. Also, we observed significant differences of mixing metrics in reactive and nonreactive systems. (C) 2010 American Institute of Chemical Engineers AIChE J, 56: 2421-2445,2010