Experiments were conducted to determine the influence of process parameters such as fill level, mixing time, shear, and baffle presence along with material attributes such as initial active aggregate size and concentration on the homogeneity of a cohesive placebo formulation in a pilot plant scale tote blender. The formulation was a ternary system made up of microcrystalline cellulose, NaCl or KCl salt, and magnesium stearate. Blend homogeneity was evaluated by sampling the blend using core samplers. Salt concentration was quantified using a conductivity technique. After a brief transient mixture, homogeneity became insensitive to mixing time and initial active concentration, however, it was a strong function of fill capacity and initial active aggregate size. Sixty percent fill was found to be optimum. Active aggregate size had an adverse effect on mixture homogeneity. The best results were obtained when the salt aggregates were initially comparable in size to that of the excipient and a high shear pre-blending step was implemented prior to dilution in the tote blender. For the strongly agglomerating material examined here, pre-blending was only beneficial if the initial aggregate size was relatively small. For cohesive systems that form large and rigid aggregates, it is recommended to mill or screen the potentially agglomerating component and then mix the system in a blender equipped with an intensifier bar.