PURPOSE: The purpose of this study was to evaluate the performance of helical CT in preoperative T-staging in patients with gastric cancer. METHOD: A total of 71 patients with an established diagnosis of gastric cancer [75 lesions, 46 early (T1) and 29 advanced (T2 or more) cancers] were evaluated with helical CT. Helical CT was performed with 5-mm slice thickness at 5-mm/s table incrementation. Using the volumetric data by helical scanning, axial CT images (5-mm slice thickness at 5-mm intervals) and multiplanar reconstruction (MPR) images were obtained. CT findings were compared with histopathologic studies of the resected specimen. RESULTS: Sensitivity of helical CT for gastric cancer was 26% (12 of 46) for early and 100% (29 of 29) for advanced cancer. Three lesions were misdiagnosed as gastric cancer by helical CT. Histopathologically, all early gastric cancers detected by helical CT were either polypoid or elevated types or showed massive invasion of the submucosal layer. The differentiation between T1 cancer with massive submucosal invasion and advanced cancer was difficult. The differentiation between T2 and T3 cancer was possible in 73% (19 of 26) and between T1/T2 and T3/T4 (extraserosal invasion) in 83% (34 of 41). Overall T-staging was correct in 66% (27 of 41). MPR images improved the detection rate (three lesions) or increased confidence in T-staging (eight lesions) over axial CT images. CONCLUSION: When helical CT detected gastric cancer that was not a polypoid or elevated type with underlying normal-appearing gastric wall, it was either T1 cancer with massive invasion of the cancer cells into the submucosal layer or advanced cancer. However, differentiation between these two stages was difficult on CT. Diagnosis of serosal invasion was not markedly improved by helical CT. MPR images increased confidence in the staging of certain gastric cancers, such as those in locations where CT images are susceptible to partial volume averaging effects.