Many pyrimidine nucleoside analogues exhibit potent anti-herpesvirus activity. Analogues of current interest in several laboratories include 5-propyl-2'-deoxyuridine, E-5-propenyl-2'-deoxyuridine, E-5-(2-bromovinyl)-2'-deoxyuridine, E-5-(2-bromovinyl)-1-β-D-arabinofuranosyluracil, 1-(2-deoxy-2-fluoro-β-D-arabinofuranosyl)-thymine (2'-fluoro-araT), 1-(2-deoxy-2-fluoro-β-D-arabinofuranosyl)-5-methylcytosine, and 1-(2-deoxy-2-fluoro-β-D-arabinofuranosyl)-5-iodocytosine. To aid in establishing the mechanisms of action and basis for selectivities of these seven analogues, the 5'-triphosphates were prepared for testing with DNA polymerases; a general method for the direct chemical synthesis of nucleoside triphosphate from nucleoside is described. The effects of the analogue triphosphates were evaluated on the following four isolated DNA polymerases: virus-induced DNA polymerases from herpes simplex virus Type 1 (HSV-1) and Type 2 (HSV-2) infections, and human DNA polymerases α and β, using conditions optimal for each. Compounds were evaluated for (a) competitive inhibition with regard to both dTTP and dCTP as independently competing substrates; (b) ability to support DNA synthesis in the absence of normally competing substrate; and (c) the effect of analogue incubation on primer template capability of resultant DNA. Competitive inhibition results indicate that all seven analogue triphosphates (a) are good inhibitors of normal substrate utilization by DNA polymerase regardless of enzyme source, (b) have much higher apparent affinities (20- to 600-fold lower K_i) for HSV polymerases than for human polymerases, and (c) are equally inhibitory to both HSV-1 and HSV-2 DNA polymerases. For example, the apparent inhibition constant (K_i) of 2'-fluoro-araTTP was 0.048 µM for HSV-1 (K_m of dTTP = 0.14 µM), 0.060 µM for HSV-2 (K_m of dTTP = 0.18 µM), 1.2 µM for human polymerase-α (K_m of dTTP = 5.4 µM), and 18 µM for human polymerase-β (K_m of dTTP = 8.6 µM); the relative abilities of competitive inhibition (in order of decreasing binding affinity as reflected by increasing K_i) were E-5-(2-bromovinyl)-araUTP > 2'-fluoroarabinoside triphosphates > E-5-(2-bromovinyl)-dUTP > E-5-propenyl-dUTP > 5-propyl-dUTP for all polymerases except human β. The analogues varied considerably in support of DNA synthesis in the absence of normally competing substrate, again with little difference between polymerases; for example, regardless of enzyme source, 5-propyl-dUTP in the absence of dTTP resulted in 60-70% DNA synthesis relative to dTTP, whereas E-5-(2-bromovinyl)-araUTP gave little or no DNA synthesis, suppressing polymerase activity below background levels. The relative ability to support DNA synthesis was generally E-5-propenyl-dUTP ≃ dTTP > E-5-(2-bromovinyl-dUTP) > 5-propyl-dUTP » 2'-fluoro-arabinonucleoside triphosphates » E-5-(2-bromovinyl)-araUTP. Incubation of analogue triphosphates and polymerase with activated DNA suggests that, with E-5-(2-bromovinyl)-araUTP as the exception, the analogues have little effect on the subsequent ability of product DNA to serve as primer template. E-5-Propenyl-dUTP exhibited behavior markedly the most similar to dTTP throughout these studies. Some general observations concerning structure-activity relationships are discussed.