I have utilized genetically engineered mouse models of pancreatic cancer to identify a potential new therapeutic target, and to test the efficacy of a putative ras inhibitor. In the first part, I show that cathepsin B is upregulated during disease progression in the mouse pancreas, as is overall cathepsin activity. Loss of cathepsin B decreases preinvasive disease burden and imparts a significant survival benefit, with a consistent decrease in proliferation. In addition, lack of cathepsin B also decreases the burden of liver metastasis. Phospho-Erk localization appears to be affected by cathepsin B loss, which may account for the defect in proliferation. Cathepsin B null tumours also have increased active cathepsin L, which may compensate for cathepsin B in tumour progression and metastasis. Finally, a cysteine protease inhibitor in combination with gemcitabine confers a significant increase in survival in tumour-bearing mice. In the second part of this work, I have investigated the effect of farnesylthiosalicylic acid (FTS) in pancreatic tumour-bearing mice. In combination with gemcitabine, FTS significantly increases survival and decreases tumour kinetics and proliferation, and inhibits liver metastasis. Although FTS has previously been reported as a ras inhibitor, there is no evidence of modulation of ras activity or signaling in primary tumours after long-term or short-term intervention, or in liver metastases. In short, the therapeutic effects of FTS in this mouse model of pancreatic cancer do not appear to be ras-related, and the target of FTS remains to be elucidated.