Since 2'-fluoro-5-methyl-beta-L-arabinofuranosyluracil (L-FMAU) has been shown to be a potent anti-HBV agent in vitro, it was of interest to study the structure-activity relationships of related nucleosides. Thus, a series of 1-(2-deoxy-2-fluoro-beta-L-arabinofuranosyl)pyrimidine nucleosides have been synthesized and evaluated for antiviral activity against HBV in 2.2.15 cells. For this study, L-ribose was initially used as the starting material. Due to the commercial cost of L-ribose, we have developed an efficient procedure for the preparation of L-ribose derivative 6. Starting from L-xylose, 6 was obtained in an excellent total yield (70%) through the pyridinium dichromate oxidation of the 3-OH group followed by stereoselective reduction with NaBH4. It was further converted to the 1,3,5-tri-O-benzoyl-2-deoxy-2-fluoro-alpha-L-arabinofuranose (10), which was then condensed with various 5-substituted pyrimidine bases to give the nucleosides. Among the compounds synthesized, the lead compound, L-FMAU (13), exhibited the most potent anti-HBV activity (EC50 0.1 microM). None of the other uracil derivatives showed significant anti-HBV activity up to 10 microM. Among the cytosine analogues, the cytosine (27) and 5-iodocytosine (35) derivatives showed moderately potent anti-HBV activity (EC50 1.4 and 5 microM, respectively). The cytotoxicity of these nucleoside analogues has also been assessed in 2.2.15 cells as well as CEM cells. None of these compounds displayed any toxicity up to 200 microM in 2.2.15 cells. Thus, compound 13 (L-FMAU), 27, and 35 showed a selectivity of over 2000, 140, and 40, respectively.