The Aircraft Noise Working Group (ANSWr) was established by DLR, ONERA, and NASA to compare simulation tools, establish guidelines for noise prediction, and to assess uncertainties associated with the simulation. To accomplish these goals, a benchmark problem was initiated by the group. The setup is documented and initial results for the reference aircraft are discussed. The reference aircraft is a conventional tube-andwing configuration with the engines installed under the wings. The aircraft noise simulations are performed for departure and approach conditions, and the results obtained with three different system noise prediction tools are compared. At the aircraft level, the overall agreement is good between the three predictions. Peak noise levels agree within 3-4 dB. Fan and jet component predictions are generally very similar although there can be differences of up to 7 dB in some fan tone levels. The prediction of airframe components shows the most disagreement between the three methods with some differences of 6 dB for the major components and greater differences at high frequencies. There can also be differences in the frequency of the peak level and in the rank order of the airframe components. At the total airframe noise level, the differences are reduced to no more than 3-4 dB. While there is general agreement in shape characteristics of predicted ground noise isocontours, the differences between the three methods can result in more significant disagreement in the sizes of the isocontours.