Extensive clinical trials are required for registration and approval of new antidepressants in most countries including the requirement that a minimal effective dose should be determined. The rationale for this requirement is to avoid the use of unnecessarily high doses. The implication is that for every antidepressant, a dose exists that serves as a threshold, below which all doses are not effective or are clearly less effective in treating a major depressive episode. Dose titration and fixed dose studies are used to determine the minimal effective dose, but both strategies have limitations and often do not allow definite establishment of a clear-cut minimal effective dose. The effort of establishing a minimal effective dose for registration is expensive and time-consuming, and may not be beneficial. Variations in pharmacokinetics and other important factors are likely to confound the results of studies and impair the determination of efficacy. The consequence of these confounding variables may be the need for larger patient samples in clinical trials or separate studies with sufficient sample sizes controlling for these variables. However, this raises issues of practicality and cost. For several newer and most older antidepressants, a minimal effective dose has not been established. Nevertheless, most antidepressants in clinical practice are safe and well tolerated. Methodologically sound establishment of a minimal effective dose raises ethical problems, as large numbers of patients need to be treated with expectedly ineffective doses. From several methodological, clinical and ethical considerations, it would seem unnecessary to establish a minimal effective dose for an effective and well-tolerated new antidepressant.