In this study, diffusing wave spectroscopy (DWS) is used to investigate the effect of shear on a food-related aggregating emulsion. The principle of the method is validated using a nonaggregating, nearly monodisperse latex suspension. In general, with increasing shear rate the diffusive motion of the scatterers becomes negligible compared to the convective motion. This causes a decrease in the decay time of the autocorrelation curves and a change in the form of the autocorrelation curves from nearly exponential to Gaussian. This is reflected in the exponent of the mean square displacement that changes from 1 to 2. The effect of shear on the acidification of a sodium caseinate-stabilized emulsion was studied by DWS and by rheometry. The emulsion droplets in the food-related emulsion were uniformly dispersed at neutral pH. Upon acidification down to a pH of 5.2 ± 0.05, the emulsion showed Newtonian behavior with constant viscosity over the whole pH range. At pH 5.17 ± 0.05, independent of the applied shear rate during acidification, the viscosity suddenly increased. From this point on, the emulsion showed shear-thinning behavior. The photon-transport mean free path (l*) was not influenced by the applied shear rate and did not change down to pH 5.2 ± 0.05. Close to this pH, l* increased, and the decay of the autocorrelation curves shifted to longer correlation times when shear rates smaller than 1 s−1 were applied. At lower pH (5.05 ± 0.05),l* started to fluctuate, and the autocorrelation curves no longer decayed to zero, indicating that at these shear rates the system behaved nonergodicly. Assuming that the convective motion and the Brownian motion are independent of each other, the mean square displacement as a result of Brownian motion was determined. From this, the sol−gel point and the radius of the aggregates at this point as a function of the shear rate was determined. The results indicated that the radius of the aggregates at the sol−gel transition decreased with increasing shear rate and suggested that shear will result in a more open structure of the network formed by the aggregates.