A valid measure of dependence on cigarettes is a useful tool for clinicians and researchers. The aim of this study was to develop a new, self-administered measure of cigarette dependence, and to assess its validity. The content of the instrument was generated in qualitative surveys. A long version (114 items) was tested on the internet in 3009 smokers. Subsamples provided retest data after 18 days (n=578), follow-up data after 45 days (n=990) and saliva cotinine (n=105). The study resulted in a 12-item scale labelled the Cigarette Dependence Scale (CDS-12), and in a 5-item version of this scale (CDS-5). Except for tolerance, CDS-12 covers the main components of DSM-IV and ICD-10 definitions of dependence: compulsion, withdrawal symptoms, loss of control, time allocation, neglect of other activities, and persistence despite harm. CDS-5 has similar measurement properties but less comprehensive content. Both scales had a high test-retest reliability (r>or=0.83), and a high internal consistency (Cronbach's alpha>or=0.84). CDS-12 scores were higher in daily smokers than in occasional smokers (+1.3SD units), and were associated with the strength of the urge to smoke during the last quit attempt (R(2)>or=0.25), and with saliva cotinine (R(2)>or=0.17). CDS-12 and CDS-5 scores decreased in daily smokers who switched to occasional smoking at 18-day retest. Dependence scores did not predict smoking abstinence at follow-up. In conclusion, CDS-12 and CDS-5 are reliable measures of cigarette dependence which fulfill several criteria of content validity and construct validity and are sensitive to change over time.