This paper describes the artificial  neural  network application to short term load forecasting of an electrical utility. Load forecasting plays an important role in power system operation, planning and control. It has long been recognized that accurate short term load forecast represents a great savings potential for electric utility corporations. Various approaches like time series, regression, expert systems and artificial neural networks have been envisaged in power system operation. A case study using  ANN based load forecasting was developed. The capability of Back Propagation algorithm and Kohonen Network have been applied for load forecasting. The performance of the above two methods is tested with the data obtained from Tamilnadu Electricity Board. The design procedure are demonstrated and a sample results are presented.