Synthetic seismograms for laterally inhomogeneous velocity and Q structures can be calculated by the disk ray theory (DRT) method. Ray tracing through the model provides travel-time, distance, slowness, and ray amplitude data which are combined to synthesize seismic record sections by DRT. The amplitude factors include geometrical spreading, attenuation due to Q-1, reflection and transmission coefficients, and the free surface conversion coefficient. Comparison of DRT synthetic seismograms with those calculated by the reflectivity method for a layer over a half-space model shows excellent agreement in both amplitude and waveform character of all arrivals. Direct, reflected, P to S converted, multiply reflected, and head waves may be synthesized by the DRT method. In an application of two-dimensional synthetic seismogram modeling using the DRT method, synthetic seismogram record sections were calculated for a complicated geological model of the crust in the eastern Snake River Plain, Idaho and compared with observed seismic data. In this application, we find that a small modification of the Snake River Plain velocity model which was interpreted from travel-time data produces DRT synthetic seismograms which compare well with the amplitude and waveform character of the observed seismograms.