Rapid growth in electricity network peakdemand is increasing pressure for new investment which may be used for only a few hours a year. Residential air conditioningis widely believed to be the prime cause of therise in peak demand but, in the absence of detailed residential demand research, there is no bottom-up empirical evidenceto support this supposition or to estimate its impact. This paper first examines the developments in network peakdemand, at a national, network distribution, and local distribution feeder level to show recent trends in peak demand.Secondly, this paper applies analytics to the half-hourlyconsumption data of a sample of Ausgrid’s interval meteredcustomers, combined with local weather data, to develop analgorithm which can recognize air-conditioner use and canidentify consumption patterns and peak load. This estimate isthen compared to system peaks to determine residential airconditioning’simpact on overall demand. Finally, this paper considers the future impacts of air-conditioning load on peakdemand as penetration rates reaches saturation levels andnew minimum energy performance standards take effectreducing new units peak impacts.