The terrestrial noise experiment on the Ariel III satellite is designed to measure the radio noise from lighting discharges, and to deduce the distribution of the sources. The measurements are made at high frequencies; at optimum frequencies the noise penetrates the ionosphere only at near vertical incidence and the sources can therefore be localized. The receivers operate in narrow bandwidths in the standard frequency bands, to minimize interference from transmitters, Measured parameters are the average field strength of the noise and the numbers of atmospherics with amplitudes exceeding a predetermined value. Interpretation of the data requires a knowledge of ionospheric characteristics, which are derived from both ground stations and satellite experiments. Results so far show that noise intensities are in broad agreement with what could be expected from results of previous experiments on the ground, where such data exist, and are providing data for areas of the world where there are no noise-measuring observatories. There is some evidence that atmospherics from lighting discharges in temperate and tropical regions have significantly different radio spectra, in support of previous limited measurements at ground stations. Effects of interference from standard frequency transmissions have been studied, and some comparisons have been made between the noise data and cloud photographs from meteorological satellites.