Vane shear test data obtained by a number of researchers show that the excess pore pressures generated within the soil surrounding the vane by vane insertion and rotation and their effects on the measured vane shear strength have been misinterpreted for many years. The accepted model developed by Blight of field vane testing and the accepted criteria for determining undrained and fully drained vane shear strengths are based on this misinterpretation. Consequently, estimates that are based on this model of the degree of drainage that has been attained at the time the vane shear strength is measured may be significantly in error, and the measured undrained shear strengths may be unconservative. A revision of Blight's approximate theory of field vane testing is presented which is consistent with the available experimental data. Revised practical criteria for determining the undrained and fully drained shear strengths are also presented, and a simple revision of current standard vane shear test methods is proposed which would eliminate, for all but those soils with very high coefficients of consolidation, the possibility that estimates of the undrained vane shear strength may be unconservative.