Investigations were made into the factors such as irradiation, gap length, gas pressure and electrode conditioning which affect the time lag and the corona and breakdown characteristics in sulphur hexafluoride. The measurements of time lags were made with uniform-field electrodes, using approximately square-pulse voltages superimposed upon direct-approach voltages. Time lags in SF6 under these conditions were long, and the results were widely scattered. At low overvoltages, the time-lag distribution curves were characterised by a region free from time lags, which separated two groups of time lags. In air under the same conditions, the observed time lags were much shorter, with correspondingly smaller scatter. Additional measurements of time lags in SF were made using impulse voltages of 0.5/320μs, without the approach voltage, with uniform-field electrodes and 6.25cm-diameter spheres. The average time lags measured with the impulse voltage were of the same order as those measured with a square voltage pulse superimposed upon direct-approach voltages. Illuminating the cathode by ultraviolet light (emission current of approximately 10−11 A/cm2 caused a large reduction in the direct breakdown voltage (up to 30%) for the short gaps. This large reduction could not be explained by considering the statistical scatter observed in the breakdown voltages. It is suggested that the effect is associated with the spark mechanism in SF6 Investigations were also made of the growth of corona current, and the corona-onset voltage in SF6 and SF6-air mixtures. In SF6, large corona currents were observed over a wide range of voltage, while in air under similar conditions, corona could be detected only at voltages close to breakdown. The addition of SF6 to air caused a large increase in the corona current, the corona-onset voltage, the corona-voltage range and the breakdown voltage.