Twenty-five newly admitted patients were randomly assigned to supplemental treatment with an oral fructose mineral drink or a placebo solution. Nine pints were given in a double-blinded study during the first 3 days. Observations of tremor, chlordiazepoxide requirements, patients perceptions, and observer perceptions of common withdrawal symptoms and signs and the serum levels of Zn, Mg, and potassium were made initially and after the treatment. Although the patients improved over the 4 days of observations by many of the criteria, there were no differences between those receiving active solution and placebo.