Background <p>Smoking is a leading risk factor for numerous adverse health outcomes. However, in the context of a dramatically increasing aging population, evidence on how different dimensions of smoking characteristics is associated with biological aging remains limited and inconclusive.</p> Methods <p>This cross-sectional study involved 143,955 participants aged ≥ 65 years from the Chinese National Basic Public Health Services Project in 2021. A composite biomarkers-based biological age (BA) was constructed using the Klemera-Doubal method (KDM). We estimated the associations of different dimensions of smoking (including smoking status, smoking initiation age, cigarettes smoked per day, smoking duration, years of smoking-to-age ratio, and cumulative smoking pack-years) with the value, and the severity of KDM-BA acceleration using multiple linear regression and multinomial logistic regression, respectively. Restricted cubic splines (RCS) were used to examine their dose-response relationships.</p> Results <p>All dimensions of smoking behavior were significantly associated with accelerated KDM-BA and elevated risk of severely accelerated aging (Δ ≥ 2 years). Compared with never smokers, participants who were current smokers, smoked since childhood, smoked more than 20 cigarettes per day, smoked for more than 50 years, smoked for more than 70% of their years of life, and smoked more than 40 pack-years, had an increased KDM-BA acceleration by 0.33 (95% CI: 0.30–0.36), 0.55 (95% CI: 0.29–0.81), 0.32 (95% CI: 0.29–0.35), 0.35 (95% CI: 0.31–0.38), 0.35 (95% CI: 0.32–0.39), and 0.33 (95% CI: 0.30–0.36) years, respectively. According to results of RCS, smoking duration and years of smoking-to-age ratio showed a linearly increasing dose-response relationship with KDM-BA acceleration, while the patterns for cigarettes smoked per day and cumulative smoking pack-years exhibited a rapid increase initially, but followed a slight decline after reaching a peak at approximately 15 cigarettes/day and 55 pack-years. Among participants ever smoked, earlier smoking initiation age was associated with greater KDM-BA acceleration, reaching a plateau after age 20.</p> Conclusions <p>Different dimensions of smoking show robust and dose-response relationships with composite biomarkers-based BA acceleration. Our findings confirm that there is no safe level of smoking and emphasize the public health significance of achieving small but population-wide shifts in biological age by strengthening tobacco control measures.</p>