Treatment of pancreatic cancer involves combinations of therapeutics including use of the chemotherapy agent gemcitabine. This thesis aimed to identify gemcitabine resistance mechanisms in pancreatic cancer and so enable discovery of biomarkers to tailor chemotherapy. Clonally and adaptively resistant pancreatic cancer cell lines were developed, producing five clonally resistant cell lines (Suit-2R, Suit-2R2, Suit-2R3, Suit-2R4, Panc-R), and an adapted line (Suit-2G+). Removing the drug from Suit-2G+ restored sensitivity, giving Suit-2G- cells. Gemcitabine can cause double strand DNA (dsDNA) breaks by incorporation into DNA and as a result of replicative stress induced by deoxynucleotide (dNTP) starvation. dsDNA breaks were assessed with neutral comet assays, using gemcitabine or gamma radiation. All resistant cell lines had less damage than their parent cell line did at low doses of gemcitabine; Suit-2R2 had increased dsDNA breaks at high (250nM) levels. Suit-2G cells were less sensitive to radiation and for Suit2-R2, R3 and R4 it was greater, with the greatest increase in R4. Gemcitabine starves cells of dNTPs by inhibiting ribonucleotide reductase (RR). Suit-2R and Suit-2R4 had low levels of the RR subunit RRM1, Suit-2R2, Suit-2R3, Panc-R and Suit2G+ had higher levels. Replicative stress caused by gemcitabine incorporation, due to dNTP starvation or dsDNA breaks, can be prevented by checkpoint arrest. Expression of the checkpoint protein p21 was reduced in all resistant Suit-2 lines but increased in Panc-R. Chk1 was reduced in Suit-2R and increased in Suit-2R3, R4 and Panc-R. Suggesting that in Suit-2R there was a reduction in need/ability to activate checkpoints, while in Suit-2R3 and R4 checkpoint arrest had become predominantly Chk1 dependent (more G2/M focused). Replicative stress can also be reduced by decreasing Cyclin D1 or by relative increase in the D1b isoform (slowing S-phase entry). Panc-1 is homozygous for G870A of cyclin D1 and so has more D1b, whereas Suit-2 is heterozygous. Suit-2G+/G- had increased D1a. Suit-2R and Suit-2R2 have reduced D1 with a higher proportion of D1b. Suit-2R3 and Suit-2R4 have higher D1a. Cell cycle analysis tied together these results and suggested distinct resistance mechanisms. A: Suit-2G+ has a normal cell cycle; increased nucleotide pools overcome replicative stress. B: Suit-2R, Suit-2R2 pause at G1/S, preventing S-phase entry. Starved nucleotide pools in Suit-2R, and low Cyclin D1 and Chk1 inducing an arrest in Suit-2R2. C: Suit-2R3/R4 have many viable cells in G2, driving through S phase accepting replicative stress. D: Panc-R has a measured and controlled cell cycle through high RRM1. The different ways of achieving resistance identified in cell lines may have equivalents in patients, identification of the dominant mechanism can be exploited as a therapeutic biomarker.