MODERN theories of direct-stroke protection premise that the ground wires are so located as to intercept the stroke and provide perfect shielding. In spite of the fundamental importance of this question there still exists considerable doubt as to the correct position of the ground wires relative to the transmission conductors. Two avenues of approach are suggested for the attack of this question; first, the collection of statistical information regarding actual line performance, and second, the use of laboratory models. Line performance is, after all, the final criterion. However, it is difficult to isolate the shielding effect from other factors which may produce outage due to lightning. Studies with models eliminate this difficulty but always contain the element of doubt as to whether the laboratory conditions are sufficiently representative of actuality as to justify general conclusions. Perhaps the best course, that which is attempted in this paper, is to try both and co-ordinate the results so obtained.