The sine wave test is maybe the most important method for characterizing ADC's. By this, the acquisition device is excited with a sinusoidal signal, and a long series of output values is measured. With the help of these observations, the parameters of the DUT can be determined. The general method to do this is the Least Squares (LS). In this paper, we present a similar method using the Maximum Likelihood Estimation (MLE). It is more robust than the LS method, which has nice properties only under special conditions. This maximum likelihood problem is solvable only numerically. For this, a numerical method is presented, and simulation results are given. The main message of this paper is how to handle the problems of the estimation in the best way in order to extract possibly the full information from the measured data, and obtain a robust, effective algorithm.