Results of a study of a Beluga trained to discriminate between two complex planar targets, indicate that this animal's echolocation discrimination capability with and without blindfolds is superior to that of bottlenosed dolphins tested on similar targets. The standard target used was a three‐step pyramid constructed of three polyvinyl chloride sheets each 13 mm thick. The first step was 10×10 cm, the second step was 7×7 cm and the third step was 3×3 cm. The selection of the standard target was considered a correct response. The comparison targets (incorrect choice) differed from the standard target only in the size (surface area) of the third step. With a comparison of a third step of 2.9×2.9 cm versus the standard, 3.0×3. 0 cm, the animal's mean correct response was 80.8% correct. When presented with two standard or identical targets the mean response dropped to chance (55. 5%). The performance of the bottlenosed dolphins tested on similar targets dropped to chance on a comparison of 2.7×2.7 cm versus 3.0×3.0 cm. The pulses used by the Beluga during the echolocation discrimination task contain very little acoustic energy below 12 kHz with the peak energy occurring at 40, 80, and 120 kHz. The measurements of the frequency components in the echolocation signal above 110 kHz are only tentatively believed due to limitations in the frequency response characteristics of the hydrophones used. The ability to use both sensory channels (tests without blindfolds) appeared to enhance discrimination ability slightly. [Work supported by NUC.]