To the Editor.—In the course of a recent reanalysis of the clinical database of the Stanford (Calif) Mental Health Clinical Research Center, errors were encountered in the descriptions of certain research subjects (admitted to the Clinical Research Center between 1975 and 1980) and procedures that appeared in several published reports from the Center. One of these articles was published in the Archives.1 Among the current uncertainties are the classification of certain control subjects and the possibility that some of the subjects were not drug free. Because of the various uncertainties, we feel we must retract the article. Based on other literature, we still feel that our original hypotheses of lowered cerebrospinal fluid homovanillic acid in Huntington's disease and correlations between cerebrospinal fluid homovanillic acid concentrations and severity and duration of illness are tenable, but we cannot test these hypotheses with the data reported in the above-mentioned article. Until further experiments are done, low concentrations of cerebrospinal fluid homovanillic acid cannot be used as a putative marker for early Huntington's disease. We apologize for any confusion this situation may have caused.