In 227 smokers' clinic clients who managed at least one week of abstinence, ratings of withdrawal symptoms were used to predict subsequent return to smoking. Depression, time spent with urges to smoke, and difficulty not smoking during the first week of abstinence were significantly associated with lapse back to smoking during the second week. Depression and difficulty not smoking during the second week significantly predicted smoking status during the third week. No withdrawal symptoms during the third week were associated with a return to smoking during the fourth week. The predictive power of withdrawal symptoms was over and above that of smoke intake and dependence. The findings provide some support for the view that reduction of withdrawal symptoms may be a worthwhile target in programmes for helping smokers to quit.