The paper gives an analysis of the 1157 faults due to lightning which have occurred during 14 years on the 5 900 miles of line of the British Grid system.It is shown that the fault rate varies approximately inversely with the system voltage, typical figures being 1 fault per 100 miles per year for the 132-kV lines, and 3.8 for the 33-kV lines. The annual, seasonal and diurnal variations in the incidence of lightning faults are studied, and figures are given for the mean fault rates in the different Grid areas. Factors of importance in the design of overhead systems are discussed, such as the reliability of double-circuit lines, the proportion of single-phase to multi-phase flashovers, and the effect of tower footing resistance. It is found that overhead lines can be put back into service again immediately after a lightning faultin 9 cases out of 10.It is estimated that, for an average Grid substation, there is not more than one fault due to a direct stroke in about 70 years. The ecords of transformer breakdowns are discussed in some detail, and it is shown that the most serious form of breakdown, failure of windings, has involved, on theaverage, 0.3 transformer per 100 transformers per year.The protective devices used on the Grid system are described, and information is included on the performance in service of rod gaps, expulsion tubes and Petersen coils. The rod gaps and expulsion tubes appear to have provided a useful measure of protection, and the 33-kV system fault data lead to the conclusion that the Petersen coils suppress about 50% of the faults. Some information, based on operating experience, is given on the protective effect of terminal cables.Finally, the use of radio methods for giving thunderstorm warnings and statistical information on lightning severity is briefly discussed.