Chemically modified oligonucleotides are being developed as a new class of medicines for curing conditions that previously remained untreatable. Three primary classes of therapeutic oligonucleotides are single-stranded antisense oligonucleotides (ASOs), double stranded small interfering RNAs (siRNAs), and oligonucleotides that induce exon skipping. Recently, ASOs, siRNAs, and exon skipping oligonucleotides have been approved for patients with unmet medical needs, and many other candidates are being tested in late stage clinical trials. In coming years, therapeutic oligonucleotides may match the promise of small molecules and antibodies. Interestingly, in the 1980s when we developed chemical methods for synthesizing oligonucleotides, no one would have imagined that these highly charged macromolecules could become future medicines. Indeed, the anionic nature and poor metabolic stability of the natural phosphodiester backbone provided a major challenge for the use of oligonucleotides as therapeutic drugs. Thus, chemical modifications of oligonucleotides were essential in order to improve their pharmacokinetic properties. Keeping this view in mind, my laboratory has developed a series of novel oligonucleotides where one or both nonbridging oxygens in the phosphodiester backbone are replaced with an atom or molecule that introduces molecular properties that enhance biological activity. We followed two complementary approaches. One was the use of phosphoramidites that could act directly as synthons for the solid phase synthesis of oligonucleotide analogues. This approach sometimes was not feasible due to instability of various synthons toward the reagents used during synthesis of oligonucleotides. Therefore, using a complementary approach, we developed phosphoramidite synthons that can be incorporated into oligonucleotides with minimum changes in the solid phase DNA synthesis protocols but contain a handle for generating appropriate analogues postsynthetically.This Account summarizes our efforts toward preparing these types of analogues over the past three decades and discusses synthesis and properties of backbone modified oligonucleotides that originated from the Caruthers' laboratory. For example, by replacing one of the internucleotide oxygens with an acetate group, we obtained so-called phosphonoacetate oligonucleotides that were stable to nucleases and, when delivered as esters, entered into cells unaided. Alternatively oligonucleotides bearing borane phosphonate linkages were found to be RNase H active and compatible with the endogenous RNA induced silencing complex (RISC). Oligonucleotides containing an alkyne group directly linked to phosphorus in the backbone were prepared as well and used to attach molecules such as amino acids and peptides.