The histoculture drug-response assay (HDRA) was recently evaluated in a retrospective clinical trial and was found to correlate to drug sensitivity, resistance, and patient survival. To further investigate the potential of HDRA to contribute to patient survival, 215 patients with gastric cancer from 45 medical centers were tested with the HDRA in a blinded study after resection of the primary lesion. One hundred sixty-eight patients received at least 20 mg/m2 of mitomycin C and a minimum of 30 g UFT, a mixture of tegafur and uracil at a molar ratio of 1:4, thereby making them eligible for the study. Of these cases 128 were evaluable by the HDRA. The evaluable patient tumors were tested by the HDRA with the [3H]thymidine incorporation end point measured by microautoradiography to be drug "sensitive" or "resistant." The in vitro conditions for distinguishing sensitivity and resistance that matched the response rates for historical controls for gastric carcinoma were 90% inhibition rate and 0.12 microgram/ml for mitomycin C and 70% inhibition rate and 1 microgram/ml for 5-fluorouracil, respectively. Most importantly in the blinded study, the overall and disease-free survival rates of the HDRA-sensitive group were found to be significantly higher than those of the HDRA-resistant group tested under the above conditions. The data further indicate the importance of three-dimensional tumor culture for obtaining accurate clinical information. The results demonstrate that the HDRA response correlates to patient survival, which suggests the potential of the HDRA to contribute to patient survival in gastric cancer when used prospectively.