Résumé : Background: A number of in vitro assays are used to determine susceptibility of HSV to antiviral agents, but results from these in vitro assays do not necessarily correlate with treatment outcome. Objectives: A method with improved capability for identifying an isolate as acyclovir (ACV) or penciclovir (PCV) resistant when resistance is borderline could greatly improve the management of HSV disease. Study design: A comparative evaluation of four in vitro assays, plaque reduction (PRA), DNA hybridization, plating efficiency (PEA) and plaque autoradiography (PAR) was performed to accurately identify and measure resistance of a TK-altered clinical HSV isolate (HSV-1 N4) from a patient who was non-responsive to ACV treatment. Two established criteria for the prediction of antiviral resistance, IC 50≥2.0 μg/ml or an IC 50 greater than 10× above a sensitive virus IC 50, as well as testing in human (MRC-5) and nonhuman (Vero and CV-1 monkey kidney) cell lines were evaluated. Results: The PRA and DNA hybridization assays accurately identified HSV-1 N4 as ACV r in human cells when using the 10× above sensitive virus IC 50 resistance criterion. Moreover, the PEA and PAR assays failed to classify HSV-1 N4 as drug resistant and indicate that these technologies alone are inadequate for identifying resistant virus. Conclusions: The data presented herein indicate that the PRA and DNA hybridization assays most accurately identified an otherwise borderline-resistant isolate as drug resistant: (i) when a sensitive virus is used within each individual assay as a control, (ii) when ACV and PCV susceptibility is evaluated in human cells, and (iii) when the 10× above sensitive IC 50 criterion is used to classify a virus as drug-resistant. Testing of additional clinical samples is warranted to further confirm these findings. © 2002 Elsevier Science B.V. All rights reserved.