The large velocity contrast between salt and the surrounding sediments generates strong conversions between P‐ and S‐wave energy. The resulting converted events can be noise on P‐wave migrated images and should be identified and removed to facilitate interpretation. On the other hand, they can also be used to image a salt body and its adjacent sediments when the P‐wave image is inadequate. The converted waves with smaller reflection and transmission angles and much larger critical angles generate substantially different illumination than does the P‐wave. In areas where time migration is valid, the ratio between salt thickness in time and the time interval between the P‐wave and the converted‐wave salt base on a time‐migrated image is about 2.6 or 1.3, depending upon whether the seismic wave propagates along one or both of the downgoing and upcoming raypaths in salt as the S‐wave, respectively. These ratios can be used together with forward seismic modeling and 2D prestack depth migration to identify the converted‐wave base‐of‐salt (BOS) events in time and depth and to correctly interpret the subsalt sediments. It is possible to mute converted‐wave events from prestack traces according to their computed arrival times. Prestack depth migration of the muted data extends the updip continuation of subsalt sedimentary beds, and improves the salt–sediment terminations in the P‐wave image. Prestack and poststack depth‐migrated examples illustrate that the P‐wave and the three modes of converted waves preferentially image different parts of the base of salt. In some areas, the P‐wave BOS can be very weak, obscured by noise, or completely absent. Converted‐wave imaging complements P‐wave imaging in delineating the BOS for velocity model building.