Twenty-seven adults with a diagnosis of the chronic fatigue syndrome were enrolled in a double-blind, placebo-controlled study of acyclovir therapy. The patients had had debilitating fatigue for an average of 6.8 years, accompanied by persisting antibodies to Epstein-Barr virus early antigens (titers greater than or equal to 1:40) or undetectable levels of antibodies to Epstein-Barr virus nuclear antigens (titers less than 1:2) or both. Each course of treatment consisted of intravenous placebo or acyclovir (500 mg per square meter of body-surface area) administered every eight hours for seven days. The same drug was then given orally for 30 days (acyclovir, 800 mg four times daily). There were six-week observation periods before, between, and after the treatments. Three patients had acyclovir-induced nephrotoxicity and were withdrawn from the study. Of the 24 patients who completed the trial, similar numbers improved with acyclovir therapy and with placebo (11 and 10, respectively). Neither acyclovir treatment nor clinical improvement correlated with alterations in laboratory findings, including titers of antibody to Epstein-Barr virus or levels of circulating immune complexes or of leukocyte 2',5'-oligoadenylate synthetase. Subjective improvement correlated with various measures of mood. We conclude that acyclovir, as used in this study, does not ameliorate the chronic fatigue syndrome. We believe that the clinical improvement observed in most patients reflected either spontaneous remission of the syndrome or a placebo effect.