Shear wave velocity measurement in the laboratory, by means of Bender Elements is becoming more and more popular for geotechnical characterization. The reasons of the increasing popularity of such method are well known and among the many the “apparent” simplicity of the testing method should be mentioned. Anyway, several researchers have pointed out the difficulties in interpreting test results. Difficulties are mainly due to the fact that real test conditions mismatch the given assumptions. The paper clearly shows, using laser measurements, that a source-bender, excited by a sine wave, oscillates non-symmetrically showing a waveform very different than the electrical input. This make questionable the possible use of cross-correlation (referring to the input waveform). On the other hand the source-bender does not bend as a cantilever when using high input frequencies. Under such circumstances, deformation mode of the BE becomes very complex and a very complex wave field is generated. Numerical simulations clearly show what above indicated. The authors conclude suggesting the use of BE only as receivers; generation of the wave motion should be accomplished with other means.