Background: Why adolescents' drinking is associated with their parents' drinking remains unclear. We examined associations in a prospective cohort study, adjusting for socio-demographic characteristics and family factors. Methods: We recruited 1927 children from grade 7 classes (mean age 13 years), and one of their parents, in three Australian states, contacted participants annually from 2010 to 2014, and analysed data from assessments at ages 13, 14, 15 and 16 years. We used the Alcohol Use Disorders Identification Test-Consumption (AUDIT-C) subscale to identify hazardous drinking in parents (score ≥5) and children (score ≥3) and constructed mixed-effect logistic regression models, accounting for clustering within school and adjusting for likely confounders. We evaluated the sensitivity of estimates by imputing missing values assuming the data were missing at random vs. missing not at random. Results: Parent hazardous drinking predicted mid-adolescent hazardous drinking, e.g. 15 years olds whose parents [adjusted odds ratio (aOR) 2.00; 95% confidence interval 1.51-2.64] or parents' partners (aOR 1.94; 1.48-2.55) were hazardous drinkers had higher odds of being hazardous drinkers at age 16. The magnitude of univariate associations changed little after adjusting for covariates, and sensitivity analyses confirmed the robustness of the association, across a wide range of assumptions about the missing data. Conclusions: The associations between parents' and their adolescent children's hazardous drinking are unlikely to be due to confounding by socio-demographic and family factors. Parents should be encouraged, and supported by public policy, to reduce their own alcohol consumption in order to reduce their children's risk of becoming hazardous drinkers.