In vitro susceptibility to acyclovir of 96 strains of herpes simplex virus isolated from 80 immunocompromised patients attended in our hospital was studied by the cytopathic effect reduction assay. Ninety-eight percent (61/62) of herpes simplex virus 1 strains and 91% (31/34) of herpes simplex virus 2 strains were inhibited by acyclovir concentrations lower than 3 mg/l. In 5% of the patients herpes simplex strains resistant to acyclovir (ID(50) >3 mg/l) were isolated. Ninety-eight percent of the lesions caused by herpes simplex viruses susceptible to acyclovir (ID(50) <3 mg/l) resolved independently of treatment. In two cases, the cytopathic effect reduction assay was not able to predict treatment failure and persistance of the lesions was not always associated with isolation of a resistant strain in vitro. In four cases, isolation of a strain resistant to acyclovir was not indicative of treatment failure. In conclusion, we believe there is no need to routinely test susceptibility of herpes simplex viruses to acyclovir and that susceptibility testing should be indicated only in patients in whom lesions persist and other causes have been ruled out.