The synthesis and the reverse transcriptase inhibitory activity of new 5'-O-mono-, di- and trifluoromethylphosphonate derivatives of nucleosides and 2'-deoxynucleosides are described. In the biosynthetic pathway leading to DNA, several enzymes using mono-, di-, and trinucleotides as substrates play a critical role. Most of the known inhibitors of these enzymatic reactions are nucleotides. However, these nucleotides are labile in vivo and are not therapeutically useful. Phosphonate esters have been used extensively to replace naturally occurring phosphates (type I, either the bridging oxygen (type II) or one of the phosphate hydroxyl functions (type III) being replaced by a carbon atom). The resulting analogs were presumed to be less susceptible to either acidic or enzymatic cleavage. The 5'-methylene, 4'- or 5'-methylenoxy (type II), 5'-methyl-, 5'-hydroxymethylphosphonate (type III) of nucleosides have been described to mimic mononucleotides, but no significant biological activity of such derivatives has been reported so far. The potential of the fluorine atom in the synthesis of nucleotide analogs has been illustrated by Blackburn et al and very recently by Hebel et al in their work on di- and triphosphates of nucleoside analogs. In both cases a fluoromethylene moiety replaces the β, γ bridging oxygen in the terminal pyrophosphate function in order to obtain an isosteric and isopolar pyrophosphate function. Fluorine atom incorporation into organic molecules has often been associated with profound changes in their biological profile compared to their non-fluorinated counterparts. Such differences are the consequences of the extreme electronegativity of the fluorine atom and its versatility in replacing either a hydrogen atom without notable steric consequences, or a hydroxyl group, the fluorine atom being able to form hydrogen bonds. In order to combine these isosteric and isopolar effects of the fluorine atom, we propose mono-, di-, and trifluoromethylphosphonate functions of type III as "bioisomers" of the phosphate group in mononucleotides. In this manner, the fluorinated moiety could mimic the missing hydroxyl function. Recently, an oligonucleotide containing a difluoromethylenephosphonate diester moiety was synthesized for 19F NMR studies, but no biological data has been reported yet. We describe here a general and straightforward route to this class of compounds, starting from the appropriately protected nucleosides or 2'-deoxynucleosides and from mono-, di-, or trifluoromethylphosphonic acids (Scheme 1, Table 1 or Scheme 2, Table 2, respectively). The general procedure for their syntheses is based on that described by Myers et al: the protected nucleoside or 2'-deoxynucleoside and fluoromethylphosphonic acid are processed in anhydrous pyridine with dicyclohexylcarbodiimide, followed by hydrolysis, deprotection, and purification via ion exchange chromatography to yield the product as a white hygroscopic solid (sodium salt for long-term storage). The chemical stability of these pseudonucleotides was evaluated: for example, the 5'-fluoromethylphosphonate of AZT showed no significant hydrolysis at pH 7.35 (50°C, 12 days) and only slight hydrolysis at pH 1.15. Evaluation as substrates/inhibitors of cellular kinases (GMP, AMP, TMP kinases) showed no activity at concentrations above Km values. Despite not being triphosphate analogs, the compounds were tested against avian myeloblastosis virus (AMV) and HIV-1 reverse transcriptases (RT). AZTTP was a potent inhibitor (as expected), while AZT/AZTMP were inactive. Notably, several fluorophosphonates exhibited RT inhibition: AZT phosphonates had the best profile, with the trifluorophosphonate derivative 11 being most potent. The mechanism is under investigation. These compounds do not require phosphorylation for activity, offering new perspectives for reverse transcriptase inhibitor design.