Although cannabis is the most commonly used illicit drug, duration of cannabis use is typically short, with many of those who initiate cannabis use ceasing use by their late twenties. In this paper we analyze data from a volunteer Australian cohort of 6265 male and female twins to examine whether the duration of cannabis use is an informative phenotype for future genetic analyses. Genetic modeling indicated: (a) moderate genetic influences on duration of cannabis use in both males (41%; 95% CI=31-51) and females (55%; 95% CI=46-63); (b) strong genetic influences on cannabis dependence in both males (72%, 95% CI=61-81) and females (62%, 95% CI=48-74); (c) no evidence of shared environmental influences on duration of cannabis use or on cannabis dependence in either males or females. Importantly, this model fitting indicated that a substantial component of genetic influences (rg=.90, 95% CI=.77-.99 (males); .70, 95% CI=.57-.83 (females)) on duration of cannabis use was shared with those influencing liability to cannabis dependence. While there were high genetic correlations in both women and men, lifetime duration of cannabis may be uniquely informative in assessing components of liability to cannabis use.