Objective: The efficacy of cognition-focused interventions (CFIs) for the treatment of Alzheimer’s disease (AD) has been questioned recently. To date, the specific effects of cognitive rehabilitation (CR), cognitive training (CT), and cognitive stimulation [CS] have not been analyzed due to inconsistencies in the use of the comparison groups. This work aims to analyze the differential effects of CFIs by removing the influence of the comparison group from the estimates of the effects. Method: a literature search performed in Pubmed, Proquest, and Embase databases yielded 65 potential studies, of which 33 studies with a sample size of 1,225 individuals were meta-analyzed. Each intervention group was treated as the unit of analysis to remove the confounding effects of the comparison condition. Measures of general cognitive functioning, memory and functional outcomes were compared using the hierarchical robust variance estimator metaregression. Age, education, sex, risk of bias, sample size, duration of intervention, the proportion of drop-outs, pharmacological treatment, and severity of disease were included as covariates. Results: Only CT differed from no cognition-focused interventions (NCFI) for memory outcomes in univariate analyses, but differences became nonsignificant when covariates were included in the model. CR showed a significantly higher effect in outcomes measuring functioning in targeted domains with no differences in standard cognitive tests relative to NCFI. Conclusions: This work supports previous findings questioning the efficacy of CT or CS for AD. Moving toward CFIs focused on relevant goals and including measures related to the skills, abilities or activities that are the focus of the intervention is encouraged.