To test the inverse beam stack multiple suppression method I have prepared three models of synthetic data composed of primaries and multiples. These models are displayed in figure . Each model consists of two primaries, one at 2.0s having constant amplitude and the other at 1.5s having a polarity reversal. Both primaries have a zero offset RMS velocity of 2.1km/s. The primaries in model2 have non-hyperbolic moveout which is due to a velocity function that is proportional to the cube of the offset. The far offset velocity of the primaries in model2 are 1.15 times higher than the near offset RMS velocity. The primaries generated in model2 might be expected to occur in the presence of a salt body. Superimposed on these primaries are multiple events. Two multiple events are superimposed on the primary events in model1 and model2 and occur at 1.4s and 1.92s. The multiples in model1 and model2 have RMS velocities of 1.5km/s and have amplitudes that are six times the zero offset amplitudes of the primary events. The multiples in model3 are made up of two periodic trains of events, one train with RMS velocity of 1.5km/s and the other train with RMS velocity of 1.7km/s. The period of the first train is 0.12s while the period of the second is 0.1s.