Despite two recent negative trials, most controlled clinical studies have found that when naltrexone is added to substance abuse treatment or counselling, significantly less heavy drinking is done by the patients who are willing to take most of the prescribed naltrexone. Naltrexone also reduces urges to drink and makes any slips back into drinking less pleasant. Therefore, naltrexone can be a useful adjunct to substance abuse counselling or rehabilitation programmes, as one of many tools that clinicians and patients use. However, beneficial effects are limited in scope. Naltrexone mostly does not increase the chance of staying completely abstinent but rather reduces the intensity or frequency of any drinking that does occur. Many alcohol-dependent individuals are medically ineligible or are unwilling to take naltrexone, many who start naltrexone do not continue with it and many who comply with it do not benefit. Compliance is greater for individuals who experience fewer adverse effects and who have stronger beliefs in the benefits of naltrexone, suggesting that clinicians can increase compliance by helping patients to manage adverse effects and by bolstering patients’ beliefs in the benefits of naltrexone. Alcohol-dependent individuals who are most likely to benefit from naltrexone seem to be those with close relatives who also had alcohol problems, or who have stronger urges to drink or who are more limited in cognitive abilities. Some individuals may benefit from a higher dose, particularly people with lower blood concentrations of the medication, and individuals who achieve good results may benefit from a longer course of treatment with naltrexone. In these ways, treatment can be targeted to increase the likelihood of beneficial outcomes with naltrexone.