Cultured cell populations derived from the refractory murine pancreatic ductal adenocarcinoma (Panc 02) were propagated in modified Eagle's minimal essential medium supplemented with 20% fetal bovine serum and had a doubling time of 19.1 +/- 4.7 hours (mean +/- SD). The Panc 02 cell populations were tested against 8 antitumor agents and exhibited different sensitivities to the agents in a 24-hour growth-inhibition assay. The concentration that inhibits the growth of the test culture by 50% relative to the growth of the control culture (IC50) in micromolars was determined for each agent. The IC50 values were: doxorubicin (ADR), 0.055; vincristine (VCR), 0.042; 5-fluorouracil, 1.92; cytarabine, 5.35; melphalan, 10.5; cisplatin, 17.0; carmustine (BCNU), 46.2; and lomustine (CCNU), 52.6. These IC50's were estimated to be pharmacologically attainable concentrations in mice. On a micromolar basis, the Panc 02 cells were the most sensitive to VCR and ADR and the least sensitive to BCNU and CCNU. By the use of a colony-forming assay and a 24-hour exposure period and the evaluation of each agent at 1/3 X IC50, 1 X IC50, and 3 X IC50, the degree of cell killing was greater than predicted on the basis of the IC50's determined in the growth-inhibition assay. The use of a 1-hour exposure period resulted in a very minimal reduction in viability of the cell populations except for BCNU and CCNU. It was concluded that the degree of cell killing was a function of drug concentration and time of exposure and that the pancreatic tumor in vivo should be sensitive to these agents, provided effective concentrations and exposure periods can be achieved at the tumor target sites.