OBJECTIVE This study used a prospective survey to examine (a) the relationships of early onset and prior consumption of cannabis, cigarette, and alcohol use with later cannabis use initiation, reinitiation, and persistence; and (b) whether the quantity or frequency of alcohol or cigarette use was more predictive of cannabis use initiation, reinitiation, or persistence. METHOD This study used data from the National Epidemiologic Survey of Alcohol and Related Conditions, Wave 1 (2001-2002) and Wave 2 (2004-2005), including 27,741 nonusers, 5,623 prior users, and 1,279 current cannabis users at baseline who were at risk of cannabis use initiation, reinitiation, and persistence, respectively, between baseline and follow-up assessments. We employed logistic regression to examine the effects of prior substance use on cannabis use initiation, reinitiation, and persistence, controlling for sociodemographics. RESULTS Frequency and early-onset status of cigarette use (≤14 years) and alcohol use (≤17 years) predicted cannabis use initiation and reinitiation but not persistence. While considering the potential effects of early onset and consumption levels of cannabis use, baseline alcohol and cigarette use quantity and frequency did not predict persistent cannabis use. Frequency, compared with quantity, of alcohol and cigarette use was more predictive of cannabis use initiation and reinitiation. CONCLUSIONS Early onset and prior experience with cannabis, cigarettes, and alcohol could have effects on later cannabis use, varying among the three at-risk populations. Different strategies are needed for preventing cannabis use initiation, reinitiation, and persistence, based on targeting early use of alcohol and cigarettes.