The theoretical mass-loading sensitivity of Love wave biosensors composed of thin layers of on ST-cut quartz is compared with measured sensitivities. The comparison presented here was prompted by earlier results in which theoretical predictions showed reasonable agreement for the mass-loading sensitivity, over a limited range of layer thicknesses, but were unable to predict the velocities of the Love waves, and could not reproduce the rapid loss of sensitivity as the layer thickness increased beyond its optimum value. In the earlier work, the theory was based on wave propagation in isotropic, non-piezoelectric, layered materials, combined with perturbation theory to predict the effect on wave velocity of a thin, solid, mass-loading layer. We therefore wished to determine whether the previous discrepancies between theory and experiment arose because of the use of isotropic theory to describe the material properties of the layers. Further theory was therefore performed, in which the anisotropic and piezoelectric nature of the layers was included. We show in this paper that the full theory gives an improved prediction of the velocities of the guided Love waves, again predicts the trend in the variation of mass-loading sensitivity with layer thickness, measured using sputtered gold as the mass-loading layer, and correctly predicts the optimum layer thickness at which maximum sensitivity occurs. However, the theory underestimates the maximum sensitivity and again does not predict the rapid decrease in sensitivity beyond this maximum. The reason for this discrepancy is still, therefore, unclear. One possible explanation, that electrical effects were partially responsible for the frequency changes recorded in the experiments, is discounted as the fuller theory shows that any such effects should be negligible for the Love waves being considered. We conclude that a theory based on a homogeneous guiding layer perfectly adhered to a piezoelectric substrate is not adequate to describe the measured sensitivities of the Love wave devices.