Over the past year, we have attempted to grow 132 different human tumor samples (78 from urological and 54 from non urological tumors) using a soft agar colony formation assay similar to that originally described by Salmon and colleagues. Formation of colonies in vitro occurred in 44 of 78 primary urological tumors (56%), including 63% (12/19) of renal cancers, 61% (25/41) of uroepithelial cancers and 42% (5/12) of testicular cancers. The effects of in vitro chemosensitivity were analysed using the inhibition of colony growth (more than 70%) at two different cut-off doses which were equilibrated with the achievable AUC doses for the higher cut-off point and to one-tenth of AUC for the lower cut-off point. Five of 13 (38%) drugs showed an effective rate between 10% and 38% for the lower cut-off point in vitro. On the other hand, eleven of 13 (85%) drugs. Showed an effective rate between 20% and 67% for the higher cut-off point in vitro. In the tested uroepithelial cancers, none of seven for the lower cut-off and three of seven for the higher cut-off were demonstrated to be sensitive cases from the dose corresponding colony inhibition curve for cisplatinum. Nine of the renal cancers were demonstrated for the dose corresponding colony inhibition curve for Interferon. To predict clinical correlation, 16 patients (20 drugs) were treated with identical drugs which were estimated from in vitro chemosensitivity testing. The predictability results were 75%-100% true positive and 85%-100% true negative, with 87% overall predictability. This assay can therefore be used to study differences of biological character including drug sensitivity.