Most published studies of the efficacy of naltrexone for alcohol treatment have focused on daily medication for relapse prevention among abstinent alcoholics. The present study compared the effects of naltrexone with those of placebo in a sample of early problem drinkers who received study medication either daily or targeted to situations identified by the patients as being high risk for heavy drinking. Patients (n = 153; 58% male) were randomly assigned to receive naltrexone (50 mg) or placebo on a daily or targeted basis, yielding comparable numbers of patients in each of four treatment groups. Patients were trained to use structured nightly diaries in which they recorded their alcohol consumption and medication intake. Analysis was conducted with hierarchical linear modeling. Irrespective of whether they received naltrexone or placebo, patients in the targeted condition showed a reduced likelihood of any drinking. There was a reduced likelihood of heavy drinking, both for patients who received naltrexone and for patients who were in the targeted groups (either naltrexone or placebo), although these effects diminished as the number of tablets available to the targeted groups was reduced over the 8-week treatment period. Although the effect was a modest one, daily naltrexone reduced the risk of heavy drinking in this patient group. Furthermore, use of a targeted approach to medication treatment appears to be a useful strategy for reducing both drinking and heavy drinking. Efforts to replicate these findings are warranted, since they suggest that schedules of medication administration other than daily should be evaluated for treatment of problem drinking.