We have examined the effect of superficial velocity, tracer injection location and measurement location on the mixing time in a bubble column. It was found that for the range of superficial velocities examined (0.07-0.29 m/s) there was little change in mixing time with superficial velocity. In contrast, it was found that the tracer injection point and measurement location had a large impact on the measured mixing time, with measured values ranging between 3 s and 25 s. Additionally, it was found that it was possible to introduce the tracer in such a way as to produce zones of poor mixing, leading to a high local tracer concentration. Such a finding is important in the design of bubble columns used in the bio-processing industry as these conditions can lead to a reduction in the yield of industrial bioprocesses. The second half of this work examines the use of Computational Fluid Dynamics (CFD) as a tool to model mixing in bubble columns. It was found that a computationally efficient CFD model correctly predicted the trends in mixing time for all superficial velocities and tracer addition locations examined, with the model being able to correctly predict the mixing time within the margin of variation. (C) 2014 Elsevier B.V. All rights reserved.