The acoustic properties of sea bed sediments containing occluded gas are dominated by the volume of gas contained in bubbles, the size of bubbles, and the elastic properties of the soil matrix. This study evaluated current theory developed by Anderson and Hampton to determine the sound speed and resonance frequency of gassy soils, and the models they used to determine the elastic properties of the soils. It compared calculated sound speeds, based on material properties simulated by the models, with measured sound speeds on "large bubble" laboratory soils produced in a similar manner to natural sea bed gassy soils. There was some evidence that the Anderson and Hampton equations accurately predicted sound speed at lower frequencies of bubbles resonance and below, but results were sensitive to inappropriate values for the elastic and damping properties of the soil. The bounds of sound speed based on the elastic properties of models that simulate "compressible fluid" or "suspension" behavior were grossly misleading when applied to large bubble soils. Conversely, sound speed based on models that correctly simulate the "bulk" or "matrix" properties of large bubble soils, at strain magnitudes and strain rates equivalent to acoustic signals, agreed well with measured data.