The results of the Multiple Risk Factor Intervention Trial showed no significant difference in mortality from coronary heart disease between intervention and control groups despite an apparent success of the intervention against cigarette smoking. A reanalysis of the published data indicates that the effectiveness of the smoking intervention may have been overestimated. The researchers counted those who smoked pipes or cigars at screen as nonsmokers and attempted to classify cigarette smokers who switched to pipes or cigars during the trial as successes whether or not they continued to inhale the smoke. It is unreasonable to assume a priori that inhaling pipe or cigar smoke is less dangerous than cigarette smoke. Moreover, attempting to include pipe and cigar smokers, who may have been inhaling smoke, among the successes makes biochemical validation of claims of abstinence overly lenient. When levels of serum thiocyanate are used to obtain an objective index of smoking reduction, it appears that the effect of intervention was considerably smaller than has been reported, and failed to reach design goals over the first 4 years. This could help to explain the disappointing mortality results.