The 6-day subrenal capsule assay (SRCA) chemosensitivity prediction test using fresh human tumor xenografts was performed in BALBc mice for 80 advanced cancer bearing patients. Among 97 SRCA, nine were non-evaluable because patients either received no chemotherapy or non-assayed drugs, and 12 were non-interpretable because of inadequate control growth. One hundred and six correlations were established between test results and clinical response, 56 retrospective (chemotherapy then test) and 50 prospective (test then chemotherapy). Among the 56 retrospective correlations there are 45 true negative (-/-) and three false positive (+/-) corresponding to 48 poor clinical responders; and eight true positive (+/+) and 0 false negative (-/+) corresponding to eight good clinical responders. Among the 50 prospective correlations there are 26 true negative (-/-) and four false positive (+/-) corresponding to 30 poor clinical responders; and 19 true positive (+/+) and one false negative (-/+) corresponding to 20 good clinical responders. In total among the 106 correlations there are 98% good negative correlations and 79% good positive correlations. The SRCA is practicable and is available for routine clinical use providing results relevant to cancer site and reflecting previous treatment status.