A simple ground clutter model has been analysed using nonlinear methods of power spectral estimation and the performance of these methods has been studied. The methods considered are the maximum entropy method (MEM), the maximum likelihood method (MLM), the leastsquares method (LSM), the autoregressive moving average (ARMA) method and Prony's energy spectral-density estimation (PESD). The performance of these power spectral-estimation methods has been studied for different data lengths and different model orders. © 1989, The Institution of Electrical Engineers. All rights reserved.