The aim of this study was to refine the long version of the Cigarette Dependence Questionnaire (CDQ-29) that includes 29-items by removing some redundant questions, and then examining the predictive utility of the resulting scale in terms of its ability to explain saliva cotinine level, and testing the measurement structure of the remaining items. Across three diverse independent samples of smokers, we developed a one-factor CDQ-12 that showed good internal consistency and concurrent validity that was comparable with the CDQ-29. The proposed model was a good fit for the data. The reliability coefficient was .92, and the score of the 12-item CDQ had a moderate correlation with the total score of cigarette withdrawal (r = .63, p < .01). The CDQ-12 is a reliable and valid Chinese-language instrument to assess cigarette dependence, which can be used to guide nursing interventions that support culturally and socially appropriate smoking cessation programs.