A significant number of individuals who start smoking cannabis develop a cannabis use disorder (CUD) and seek treatment on their own initiative, yet only a small percentage achieves sustained abstinence. Efficacious treatments for CUD are critically needed. We have developed placebo-controlled, laboratory procedures that target the effects of medications on discrete features of problematic cannabis use: positive subjective effects (cannabis 'liking'), reinforcement (cannabis self-administration under non-abstinent conditions), withdrawal (mood, sleep, food intake) and relapse (cannabis self-administration after a period of abstinence). Project 1 of this P50 Center refines and formalizes this model by developing a Decision Algorithm, utilizing a priori criteria to systematically define whether potential medications qualify to be tested in a clinical study, need to be tested further in the laboratory, or are rejected: (1) Positive signal: If a medication significantly decreases cannabis withdrawal and/or positive subjective effects and decreases cannabis self-administration, the medication will be tested in the clinic (Project 2), (2) Promising signal: If a medication decreases cannabis'positive subjective effects or cannabis withdrawal but does not alter cannabis self-administration, we will combine the medication with a cannabinoid agonist (nabilone) or opioid antagonist (sustained-release naltrexone: SR-NTX), respectively, and test the medication combination in the laboratory, (3) Negative signal: If a single or combination medication therapy produces no change in positive subjective effects, withdrawal, relapse or self- administration, we will halt further testing with this medication. The following medications with distinct neurobiological mechanisms will be tested: (1) Lorcaserin: 5HT2c agonists reduce behaviors reinforced by a range of drugs. Hypothesis: Lorcaserin will attenuate cannabis'positive subjective effects and facilitate abstinence initiation, i.e., reduce cannabis self-administration under non-abstinent conditions. (2) Cannabidiol (CBD): This cannabinoid has a complex neurobiology including inhibition of fatty acid amide hydrolase. Hypothesis: CBD will reduce cannabis withdrawal and relapse, (3) Doxazosin: ?1adrenergic antagonists dampen noradrenergic activity, which can result in reduced cannabis self-administration. Hypothesis: Doxazosin will attenuate cannabis'positive subjective effects and facilitate abstinence initiation, (4) Zonisamide: anticonvulsants that facilitate GABAergic while reducing glutamatergic neurotransmission reduce alcohol, cocaine and methamphetamine use in the clinic. Hypothesis: Zonisamide will attenuate cannabis' positive subjective effects and facilitate abstinence initiation. This proposal introduces a novel drug discovery algorithm to systematically screen medications for CUD, only advancing the mono- or combination therapies with the greatest likelihood of success in Project 2. Given the unique combination of laboratory and clinical expertise available in this Center, the outcome of Project 1 is predicted to both advance the field scientifically and address an unmet public health need: improving the treatment of CUD.