Until now mass media smoking cessation studies have relied almost exclusively on self-reports of smoking cessation because biochemical confirmation has not seemed practical. This study investigated the reliability of mailed salivary cotinine specimens for the determination of smoking status. Two simultaneous saliva specimens were obtained from 10 smokers and 10 nonsmokers. One of each pair was frozen immediately. The other was sent through the local U.S. mail and then subjected to additional physical agitation and heat before being frozen. All specimens were subsequently analyzed for cotinine. No cotinine was detectable in any of the nonsmokers' specimens. There was excellent correlation between the paired smokers' specimens. These results show that the mailing of saliva specimens for cotinine analysis is practical and provides accurate data on smoking status. It is an approach which could easily be used in mass media smoking cessation studies to biochemically confirm smoking behavior.