Pharmacotherapy of behavioural and psychological symptoms of dementia (BPSD) has been studied in numerous clinical trials. Their results, however, are inconclusive and do not allow simple recommendations applicable for majority of patients. Antipsychotics (including atypicals) has been proved effective for agitation and aggression, with doubtful effectiveness against delusions and hallucinations. Moreover, their efficacy is counterbalanced by safety concerns that include cerebrovascular events and related mortality. Antidepressants, particularly selective serotonin reuptake inhibitors (SSRIs), are similarly to antipsychotics effective for agitation and psychosis; this is encouraging for their wider use considering their better safety profile. Other drugs, including antiepileptics and benzodiazepines, are poorly studied to date and both their effectiveness and safety are questionable. Hierarchical clinical management of BPSD is recommended for routine practice. In this model psychotropics are allowed only after careful diagnostic process (including exclusion of delirium) and employment of non-pharmacologic interventions coupled with optimal use of cholinesterase inhibitors and/or memantine.