The treatment and/or prevention of cancer by selective down regulation of cancer causing gene (oncogene) transcription would represent a significant advance in the area of anticancer drug design. Non-canonical higher order DNA structures formed in oncogene promoter regions are novel targets for the modulation of oncogene expression. An obvious advantage of selectively targeting oncogene expression would be that general cytotoxicity would be minimized and the negative side effects of current chemotherapy approaches could be minimized or eliminated. To provide a foundation for the design of drugs that target oncogene promoter G-quadruplexes and i-Motifs, the basic understanding is required of the folding of guanine and cytosine rich sequences and how small molecules bind to these structures. The research reported here focuses on higher order DNA structures of two oncogenes, K-ras that is overexpressed in pancreatic cancer, and Bcl-2 that is overexpressed in a number of cancers, and one non-oncogene, HAR1. We have probed the overall structure, stability, and binding of a model drug compounds to G-quadruplex and i-Motif DNA structures in these genes. The overall objectives of this work were: 1) to understand the relationship between oligonucleotide sequence and intramolecular folding topology and stability, and 2) to understand the mechanisms for the selective binding of small molecules to these structures. Biophysical techniques including: microcalorimetry, spectroscopy, analytical ultracentrifugation, gel electrophoresis, and computational methods were used to characterize both the folding and the binding interactions. We have shown that the native K-ras purine and pyrimidine rich sequences form stable G-quadruplexes and i-Motifs. We have also characterized four G-rich sequences found within the reading frame of the human HAR1 gene. This is the first report on the formation of stable G-quadruplex motifs within the RF of any gene. The model drug, TMPyP4, binds to the Bcl-2, K-ras, and HAR1 G-quadruplexes by two different binding modes, end binding and intercalation. The significance of this research is that the results of the K-ras and Bcl-2 studies could lead to the design of drugs that selectively target oncogenes while the HAR1 results could provide new approaches to the treatment of Schizophrenia and Alzheimer’s disease.