Abstract Background Evidence on longitudinal associations between smoking and frailty is scarce. The objective of this study was to systematically review the literature on smoking as a predictor of frailty changes among community-dwelling middle-aged and older population. Methods A systematic search was performed using three electronic databases: MEDLINE, Embase and Scopus for studies published from 2000 through May 2015. Reference lists of relevant articles, articles shown as related citations in PubMed and articles citing the included studies in Google Scholar were also reviewed. Studies were included if they were prospective observational studies investigating smoking status as a predictor and subsequent changes in frailty, defined by validated criteria among community-dwelling general population aged 50 or older. A standardised data collection tool was used to extract data. Methodological quality was examined using the Newcastle-Ottawa Scale for cohort studies. Results A total of 1020 studies were identified and systematically reviewed for their titles, abstracts and full-text to assess their eligibilities. Five studies met inclusion criteria and were included in this review. These studies were critically reviewed and assessed for validity of their findings. Despite different methodologies and frailty criteria used, four of the five studies consistently showed baseline smoking was significantly associated with developing frailty or worsening frailty status at follow-up. Although not significant, the other study showed the same trend in male smokers. It is of note that most of the estimate measures were either unadjusted or only adjusted for a limited number of important covariates. Conclusions This systematic review provides the evidence of smoking as a predictor of worsening frailty status in community-dwelling population. Smoking cessation may potentially be beneficial for preventing or reversing frailty.