PURPOSE: To evaluate the usefulness of prone pad study in upper gastrointestinal series(UGIS) for thedetection of early gastric cancer(EGC). MATERIALS AND METHODS: During an eight-month period, 88 of 170 patientswho underwent gastrectomy due to EGC were also the subjects of prone pad study as well as double contrast(n=92),mucosal relief(n=76), or compression(n=91) studies. The EGCs were single in 84 patients and double in four. Wecompared prone pad study with the three other techniques for detecting a tumor and depicting the surroundingmucosal changes. Lesional conspicuity was rated 'complete', 'incomplete','suspicious'. or 'undetected'. Thedepiction of surrounding mucosal change was rated 'excellent', 'good', 'fair', or 'poor'. RESULTS: Mean tumorsize was 3.2cm, with a range of 0.3-9cm. Tumors were located in the antrum(n=55), angle(n=13), lower or midbody(n=16), or the sign body and cardia(n=5). Among the 92 EGCs evaluated, UGIS missed the lesion in threecases(sensitivity, 97%). The rates of 'complete' lesional conspicuity were 49% inn prone pad, 29% in compression,20% in double contrast, and 9% in mucosal relief. The rates of excellent' in depicting surrounding mucosal changewere 45% in prone pad, 11% in double contrast, 9% in mucosal relief, and 9% in compression. The tumor wasdemonstrated only in prone pad study in five(5%) of the 92 EGCs. CONCLUSION: prone pad study during UGIS improvesboth the detection rate of EGC and the depiction of mucosal change around the tumor.