The lubrication of powder/wall interfaces was investigated by tracking the evolution of the torque response during wall friction tests. A FT4 rheometer was employed to conduct this study. Two types of surfaces, borosilicate and Delrin, and 3 commercially-available pharmaceutical excipients, xylitol, microcrystalline cellulose and dicalcium phosphate dihydrate, lubricated or not with magnesium stearate, were tested. Two levels of relative humidity (RH), 20% and 60%, at which the powders were stabilized were used. The results showed that wall friction behaviors and, consequently, wall lubrication trends were strongly dependent on the nature of both the surfaces and powders. However, unexpected high data variability was obtained during the wall friction tests with non-lubricated powders sheared with a borosilicate surface at 20% RH. This variability was not observed when magnesium stearate was present in the powder formulations or when the tests were conducted at 60% RH. A particular effort was made to understand the possible causes of this variability. The results with non-lubricated MCC indicate that variability is reduced when the borosilicate surface is ionized at -20 kV prior to the wall friction tests, while with the same treatment on non-lubricated xylitol did not produce a similar effect. Thus, contrary to the so far prevailing belief in the literature, this preliminary study proves that electrostatic phenomena can, at least partially, explain the reported variability. (C) 2010 Elsevier B.V. All rights reserved.