Currently approved medications to treat cognitive deficits in Alzheimer's disease are targeting relatively late neurotransmission alterations and, thus, new approaches are needed to improve the observed modest efficacy. Employing various combinations of existing drugs is one way to achieve this aim. We can attempt to boost the clinical effects by adding drugs with different modes of action, or by adding a compound that is not neurotransmitter-based and not directly symptomatic, which may result in achieving disease progression modification. Recent studies with add-on therapies in Alzheimer's disease, apart from expanding our clinical knowledge, have also provided some new insights to guide future studies and therapeutic approaches. One such lesson is the critical necessity for precise characterization of subjects recruited to the studies, preferably with the use of biomarkers. Currently, even the best studied combination (memantine with a cholinesterase inhibitor) still requires further detailed trials to elucidate its usefulness at different disease stages, in patients with comorbidities, with or without the use of other concomitant medications.