Acute myeloid leukemia (AML) is the most deadly of the leukemias. Due to its heterogeneous genetic nature it has been difficult to find effective targeted treatments. Standard induction chemotherapy, which includes cytarabine (Ara-C) as its primary component, will in most cases result in remission, but the remission is short-lived and usually results in the presentation of an Ara-C resistant form of AML at relapse. A thorough understanding of how chemotherapy resistance develops in AML would lead to the establishment of drug profiles describing the molecular conditions under which a drug should be considered or rejected as a treatment option. With the ultimate goal of providing patient specific drug treatment options to restore or preserve chemotherapy sensitivity, the research presented here had five objectives: (1) to develop and/or evaluate mouse models of Ara-C resistance in vitro and in vivo and determine their ability to mimic the Ara-C resistance found in human disease, (2) to discover cellular mechanisms of Ara-C resistance, (3) to find effective drug partners for Ara-C in treating de novo AML, (4) to find drugs for treating chemotherapy refractory AML, and (5) to predict drug response based on the molecular profile (gene expression patterns and specific mutations) for each AML patient. Although these objectives are broad and aggressive, this research has resulted in some significant strides towards meeting them. Using gene expression microarray, transcriptome sequencing, and targeted mutagenesis via TAL endonuclease treatment, it was determined mutations in Dck were the primary factor in the development of Ara-C resistance in an in vitro model of Ara-C resistance. Drug screens were used to determine Ara-C resistant cells became more sensitive to glucocorticoids, and cladribine is an affective partner to Ara-C in treating de novo disease. Also, an in vivo model of Ara-C resistance was developed by passaging AML cells through SCID/beige mice and treating the mice with low doses of Ara-C.