A series of selective antiherpetic compounds were found to exert pronounced cytostatic activity against herpes simplex virus type 1 (HSV-1) and type 2 (HSV-2) thymidine kinase (TK) gene-transfected mammary carcinoma FM3A cells. Based on their potency and mechanism of cytostatic action, the antiherpetic compounds could be divided into two different classes. The first class encompasses (E)-5-(2-bromovinyl)-2'-deoxyuridine and structurally related analogues thereof [i.e., the cytosine derivative (E)-5-(2-bromovinyl)-2'-deoxycytidine and the 4'-thio derivative (E)-5-(2-bromovinyl)-2'-deoxy-4'-thiouridine]. These compounds are exquisitely cytostatic against FM3A/TK-/HSV-1 TK+ and FM3A/TK-/HSV-2 TK+ cells (50% inhibitory concentrations ranging from 0.047 to 0.001 microM) and inhibit tumor cell proliferation by inhibiting cellular thymidylate synthase. The second class consists of the acyclic guanosine derivatives penciclovir, buciclovir, and ganciclovir. These compounds are also more inhibitory to the HSV-1 TK or HSV-2 TK gene-transfected FM3A cells than to FM3A/0 or FM3A/TK- cells, but at concentrations that are higher than the concentrations at which the (E)-5-(2-bromovinyl)-2'-deoxyuridine derivatives proved to be inhibitory. These acyclic guanosine analogues appear to be targeted at the cellular DNA polymerase. From this study, (E)-5-(2-bromovinyl)-2'-deoxy-4'-thiouridine emerged as a promising candidate compound for the treatment of HSV-1 TK gene-transfected tumors in vivo, due to its metabolic stability (i.e., resistance to hydrolysis by thymidine phosphorylase).