Immunocompromised patients are frequently treated with guanine analogs such as acyclovir and ganciclovir. Acyclovir triphosphate, the active intracellular metabolite of acyclovir, exerts its antiviral effect by inhibiting herpesviral DNA polymerases through premature chain termination. PCR has recently been used for early detection of cytomegalovirus. However, we and others have experienced false-negative results for cytomegalovirus-PCR in patients on both acyclovir and ganciclovir. The impact of these agents on PCR assay is unknown. In an attempt to investigate the role of guanosine analogs in these false-negative results, we exposed the DNA-PCR for murine beta-actin, a murine CMV IE gene sequence, and a human CMV IEA1 product, to phosphorylated acyclovir derivatives. Varying concentrations of acyclovir-5'-triphosphate (final: 70-6000 microM) in the reaction mix resulted in an absence of detectable product at or above 490-670 microM. Inhibition was not observed with up to 1400 microM acyclovir-monophosphate. Increasing the Taq concentration to 10 units/100 microL stopped the inhibition. Our data demonstrate that acyclovir-5'-triphosphate inhibits PCR amplification of various gene products in a concentration-dependent manner. Furthermore, this inhibition appears to be specifically directed against the Taq polymerase and can be completely reversed by higher concentrations of the enzyme. Thus, false-negative PCR results for a viral gene product in patients under prophylaxis/treatment with acyclovir could potentially be due to contamination by acyclovir triphosphate. Therefore, negative PCR results in these patients need be interpreted with caution.