This paper discusses actual, often controversial, problems related to clinical investigations of antidepressants. Such problems include: inclusion and exclusion criteria (the investigated groups of patients are not representative for the population of depressed patients treated by psychiatrists), criteria of results' evaluation (50% reduction of pre-treatment number of points in the scale for measuring of depression severity is not satisfactory from a clinical point of view, this criterion could be replaced by remission), duration of treatment (4-6 weeks observation period is too short for adequate evaluation of treatment effects). The problem of using placebo is controversial from an ethical and legal point of view. This procedure could be replaced by other acceptable methodology.