The physical and chemical factors that allow DNA to perform its functions in the cell have been studied for several decades. Recent advances in the synthesis and manipulation of DNA have allowed this field to move ahead especially rapidly during the past fifteen years. One of the most common chemical approaches to the study of interactions involving DNA has been the use of DNA base analogues in which functional groups are added, deleted, blocked, or rearranged. Here we describe a different strategy, in which the polar natural DNA bases are replaced by nonpolar aromatic molecules of the same size and shape. This allows the evaluation of polar interactions (such as hydrogen bonding) with little or no interference from steric effects. We have used these nonpolar nucleoside isosteres as probes of noncovalent interactions such as DNA base pairing and protein - DNA recognition. We have found that, while base-pairing selectivity does depend on Watson - Crick hydrogen bonding in the natural pairs, it is possible to design new bases that pair selectively and stably in the absence of hydrogen bonds. In addition, studies have been carried out with DNA polymerase enzymes to investigate the importance of Watson - Crick hydrogen bonding in enzymatic DNA replication. Surprisingly, this hydrogen bonding is not necessary for efficient enzymatic synthesis of a base pair, and significant levels of selectivity can arise from steric effects alone. These results may have significant impact both on the study of basic biomolecular interactions and in the design of new, functionally active biomolecules.