Cocaine users consistently display cognitive impairments. However, it is still unknown whether these impairments are cocaine-induced and if they are reversible. Therefore, we examined the relation between changing intensity of cocaine use and the development of cognitive functioning within 1 year. The present data were collected as part of the longitudinal Zurich Cocaine Cognition Study (ZuCo(2)St). Forty-eight psychostimulant-naive controls and 57 cocaine users (19 with increased, 19 with decreased, and 19 with unchanged cocaine use) were eligible for analysis. At baseline and after a 1-year follow-up, cognitive performance was measured by a global cognitive index and four neuropsychological domains (attention, working memory, declarative memory, and executive functions), calculated from 13 parameters of a broad neuropsychological test battery. Intensity of cocaine use was objectively determined by quantitative 6-month hair toxicology at both test sessions. Substantially increased cocaine use within 1 year (mean +297%) was associated with reduced cognitive performance primarily in working memory. By contrast, decreased cocaine use (-72%) was linked to small cognitive improvements in all four domains. Importantly, users who ceased taking cocaine seemed to recover completely, attaining a cognitive performance level similar to that of the control group. However, recovery of working memory was correlated with age of onset of cocaine use-early-onset users showed hampered recovery. These longitudinal data suggest that cognitive impairment might be partially cocaine-induced but also reversible within 1 year, at least after moderate exposure. The reversibility indicates that neuroplastic adaptations underlie cognitive changes in cocaine users, which are potentially modifiable in psychotherapeutical or pharmacological interventions.