During the 1968 and 1969 lightning seasons, member companies of the Telephone Association of Canada conducted a field investigation of lightning voltages induced in outside plant toll facilities at ten selected sites across Canada. Oscillograms of longitudinal surge voltages occurring in open wire, paired, and coaxial cable were continuously photographed with an automatic camera system especially developed for the investigation during each season. On completion, approximately 10 000 useful surge photographs were obtained and analyzed. The results indicate that a standard test wave, with 1000-V peak and10 \times 1000-\mus waveshape, simulates 99.8 percent of the lightning surges encountered in paired and coaxial cables. For open wire circuits, a more suitable test wave with 2000-V peak amplitude and4 \times 200-\mus waveshape is required to simulate 99.8 percent of the lightning surges. The number of surges per storm was found to be far higher than previously assumed. The highest recorded incidence during a storm was 484 surges greater than 10 V. Surge repetition intervals were short in all three types of facility; about 15 percent were less than 2 s. The surge information presented provides fundamental data on lightning surge effects in Canadian telephone plants which were previously unavailable. The assembled data will be of value to the protection engineer when designing future protection systems, and also to the transmission engineer concerned with noise problems.