OBJECTIVES: Early age-at-onset of cigarette smoking is a known predictor of subsequent substance abuse. However, less is known about the association between age-at-initiation of regular smoking and age-at-initiation of illicit substance use. We examined the association between age-at-initiation of regular cigarette smoking with age-at-initiation of use of marijuana, cocaine, heroin & methamphetamine in the general US population. METHODS: We conducted a retrospective cohort study using NHANES data from 2005-2014, including individuals who smoked cigarettes, and had used marijuana, cocaine, heroin or methamphetamine at least once in their lifetime. Using simple linear regressions we studied associations between age-at-initiation of regular smoking and age-at-initiation of substance use, and multiple linear regressions to study association across different age categories of smoking initiation and age-at-initiation of substance use after controlling for sex, level of education, race/ethnicity, current age and income. RESULTS: In the study population (N=60936), the analysis showed that for every 12-month increase in the age-at-initiation of regular cigarette smoking, the age-at-initiation increased for marijuana by 2 months (RR-0.17, 95%CI 0.13-0.21), for cocaine by 1.7 months (RR-0.15, 95%CI 0.09-0.19) and for methamphetamine by 2.3 months (RR-0.20, 95%CI 0.09-0.29). When the age-at-initiation of smoking was 10-12 years, the age-at-initiation of substance use decreased by 2-3 years (p<0.05). In the adjusted models, age-at-initiation of smoking was still significantly associated with age-at-initiation of marijuana (RR-0.16, 95%CI 0.12-0.19), cocaine (RR-0.14, 95%CI 0.07-0.21) and methamphetamine use (RR-0.17, 95%CI 0.01-0.32). No significant association was found for age-at-initiation of heroin use in any model. CONCLUSIONS: An increase in the age-at-initiation of regular smoking is associated with an increase in the age-at-initiation of use of marijuana, cocaine and methamphetamine. Further research should explore causal pathways, and the effect of mental health and comorbidities on this association.