Experiments to test the influence of end effects on electron drift velocity measurements by the Bradbury-Nielsen time-of-flight method are described. A comparison of data taken at drift distances of 5, 10, and 50 cm in hydrogen and 5 and 10 cm in helium shows that over the EIN and pressure ranges investigated the results are independent of drift distance and that it is justifiable to consider this distance as that between the mid planes of the grids which terminate the drift chamber.The analysis of electron transport coefficient data to obtain scattering cross sections as a function of electron energy depends for its success on the availability of data of high precision (Crompton 1969; Bederson and Kieffer 1971). The transport coefficient that is subject to the least experimental error is the drift velocity, and Crompton et al. (1970) have claimed that measurements by the Bradbury-Nielsen method of the electron drift velocity in helium have an absolute error of less than 1 %.One of the assumptions made in assessing the error was that the effective drift length could be taken as the geometric distance between the mid planes of the grids used to gate the electron current. However, there are a number of effects, which we shall generally term "end effects", which can cause the effective drift distance to differ from the geometric distance, and these include contact potential differences between shutters, surface layer effects (at the top shutter), diffusive effects, the variation of shutter transmission with electron energy, the change in mean energy as the swarm of electrons traverses a shutter, and field distortion due to the sinusoidal potential applied to the shutter wires. A discussion of these effects has been given by Elford (1972).By suitable choices of experimental parameters it is possible to reduce end effects to an insignificant level over a wide range of values of EjN in most gases (E being the electric field strength and N the gas number density), although there are certain values of Ej N where some of these effects cannot be completely ignored. The influence of contact potential differences and surface layer effects can be reduced by using large applied potential differences across the drift chamber, while diffusive effects and the consequences of change in mean energy can be minimized by using high gas densities. The presence of field distortion due to potentials applied to the shutter wires can be detected by varying the magnitude of the sinusoidal potential. However, despite the choice of parameters and experimental checks there is no certainty that all