A prospective cohort study in adolescents. To evaluate whether smoking in adolescence is a risk factor of low back pain (LBP) among young adults. Smoking has been found to associate with LBP among adults. Longitudinal studies performed in adolescents are few. The children belonging to the northern Finland Birth Cohort 1986 were examined at birth and at 16 years of age and followed up by a postal questionnaire at the age of approximately 18 years. The primary outcome was LBP during the past 6-month period. Incident cases reported LBP at 18 but not at 16 years. Persistent cases reported pain at both time points. Logistic regression analysis was used to evaluate the effect of smoking exposure on any LBP in both genders separately and multinomial regression analysis was used to evaluate the effect on the severity of LBP (No LBP, "Reporting LBP," "Consultation for LBP") in the total population. Parents' socioeconomic status, physical activity, body mass index, and depressive mood were used as confounders in the analyses. Regular smoking at 16 years was associated with persistent LBP in girls (OR: 2.52; 95% CI: 1.40-4.53). Daily smoking of over 9 cigarettes at 16 was associated with persistent LBP (2.57; 1.03-6.46) and predicted incident pain in girls (2.80; 1.11-7.09). Pack-years of smoking were associated with incident and persistent LBP in the girls with an exposure-response relationship, whereas these associations were inconsistent in the boys. Pack-years of smoking at 18 years showed an exposure-response relationship with persistent Consultation for LBP, which was strongest in those with over 1.5 pack-years of exposure (5.82; 1.39-24.42). Regular smoking in adolescence was associated with LBP in young adults. Pack-years of smoking showed an exposure-response relationship among girls.