Cannabis use has been associated with impaired cognition during acute intoxication as well as in the unintoxicated state in long-term users. However, the evidence has been mixed and contested, and no systematic reviews of the literature on neuropsychological task-based measures of cognition have been conducted in an attempt to synthesize the findings. We systematically review the empirical research published in the past decade (from January 2004 to February 2015) on acute and chronic effects of cannabis and cannabinoids and on persistence or recovery after abstinence. We summarize the findings into the major categories of the cognitive domains investigated, considering sample characteristics and associations with various cannabis use parameters. Verbal learning and memory and attention are most consistently impaired by acute and chronic exposure to cannabis. Psychomotor function is most affected during acute intoxication, with some evidence for persistence in chronic users and after cessation of use. Impaired verbal memory, attention, and some executive functions may persist after prolonged abstinence, but persistence or recovery across all cognitive domains remains underresearched. Associations between poorer performance and a range of cannabis use parameters, including a younger age of onset, are frequently reported. Little further evidence has emerged for the development of tolerance to the acutely impairing effects of cannabis. Evidence for potential protection from harmful effects by cannabidiol continues to increase but is not definitive. In light of increasing trends toward legalization of cannabis, the knowledge gained from this body of research needs to be incorporated into strategies to minimize harm.