Identifying smokers at high risk of relapse could improve the effectiveness of cessation therapies. Although altered regional brain function in smokers has been reported, whether the whole-brain functional organization differs smokers with relapse vulnerability from others remains unclear. Thus, the goal of this study is to investigate the baseline functional connectivity differences between relapsers and quitters. Using resting-state fMRI, we acquired images from 57 smokers prior to quitting attempts. After 12-week treatment with varenicline, smokers were divided into relapsers (n = 36) and quitters (n = 21) (quitter: continuously abstinent for weeks 9–12). The smoking cessation outcomes were cross-validated by self-reports and expired carbon monoxide. We then used eigenvector centrality (EC) mapping to identify the functional connectivity differences between relapsers and quitters. When compared to quitters, increased EC in the right dorsolateral prefrontal cortex (DLPFC), left middle temporal gyrus (MTG) and cerebellum anterior lobe was observed in relapsers. In addition, a logistic regression analysis of EC data (with DLPFC, MTG and cerebellum included) predicted relapse with 80.7% accuracy. These findings suggest that the DLPFC, MTG and cerebellum may be important substrates of smoking relapse vulnerability. The data also suggest that relapse-vulnerable smokers can be identified before quit attempts, which could enable personalized treatment and improve smoking cessation outcomes.