ABSTRACT Despite a federal prohibition against marijuana, since 2012, recreational marijuana has been legalized in 11 states and Washington DC where over a quarter of US population live. Among these jurisdictions, ten states further opened or planned to open retail markets in near future to adults aged 21 years or older. The efforts of protecting public health in the new policy regime, however, have been complicated by the lack of knowledge regarding how individuals make purchase decisions and what regulatory measures would be effective to reduce problem marijuana use. Particularly, the rapid adoption of novel products and the coexistence of illegal markets pose unprecedented public health concerns and regulatory challenges. The overarching goal is to examine the relationships between recreational marijuana regulatory strategies and individual preferences for marijuana products. We will examine a wide range of policy measures that have potential to influence individual decisions, including those regulating product characteristics, restricting promotional features, modifying availability and context, and controlling price. We will innovatively test research hypotheses using two behavioral economics approaches with distinct yet complementary strengths, namely stated preferences approach and revealed preferences approach, and integrate them to provide calibrated estimates. Specifically, we aim: 1) To quantify the relationships of product, promotional, availability, and price attributes with individual hypothetical choices on marijuana products. 2) To quantify the relationships of product, promotional, availability, and price policies with individual real-world choices on marijuana products. 3) To correct hypothetical bias in stated preferences data with revealed preferences data. We will recruit representative adult samples of never users, former users, and current users of marijuana to complete a series of web-based surveys with rigorously developed discrete choice experiments on marijuana choices. To address concerns on hypothetical bias, we will complement these experiments with revealed preferences data on real-world choices through longitudinal cohort surveys in a subsample of the respondents. Particular attention will be given to the heterogeneities in policy impacts among never users, former users, and current users and among current users with medical, recreational, and dual purposes. The project will be the first rigorous and comprehensive investigation of the impacts of policy-relevant factors on marijuana use decisions at individual level. It will advance our understanding about the potential effectiveness of marijuana policies on choices between traditional and novel products in the emerging legal markets. It will also shed light on the unintended consequences of legal market regulations on demand in the coexisting illegal markets. Overall, the project has potential to advance the methodological framework for predicting marijuana demand...