We have observed a conspicuous example of supercritical reflection in both P- and SH-wave seismic data. Data were recorded in the Midland Basin (Texas) Project of the Conoco Shear Wave Group Shoot in 1977-1978. P- and S-wave critical angle phenomena, as observed in the data, are remarkably similar. Event amplitudes are small or undetectable at offsets out to about 2 000 ft, but at offsets from 2 500 to 3 600 ft amplitudes are higher than those of any other event. Head waves originating at the critical distance are weak but detectable. Long path multiplies of the supercritical parts of P and SH events appear at expected times and offsets. Constant velocity moveout corrections helped identify them.Sonic logs in combination with a knowledge of the lithology made it possible to model P-wave critical angle phenomena. Agreement of model results with the data was good when we assumed cylindrical wavefronts. As expected, modeling based on plane waves was unable to match observed phase and amplitude behavior.A number of potential uses for supercritical reflections in exploration and data processing readily come to mind, many of them related to the recording of relatively high amplitudes at distances where source noise is low. Observed rise in amplitude near the critical offset was very abrupt, particularly for SH-waves. This suggests that variations in the onset of high amplitudes may be useful for monitoring changes in velocity contrast at the reflecting interface.