Individualisation of an amitriptyline dose regimen offers substantial advantages over non-individualised treatment. In our study, we have compared both clinical effects, adverse effects and plasma steady-state concentrations of amitriptyline in 15 patients with major depressive disorder divided in three groups; (i) patients in group A were taking non-individualised doses of amitriptyline; (ii) patients in group B were taking doses of amitriptyline individualised by modified Bayesian method; and (3) patients in group C were taking doses of amitriptyline individualised by the multiple point method. The treatment course was 8 weeks long, in the setting of a psychiatric clinic. The patients in group A were taking significantly higher doses throughout the treatment course; the initial doses for the patients in group B were higher than doses for the patients in group C, but after corrections based on measured steady-state plasma concentrations they became similar. While Hamilton score descended uniformly in all three groups, both adverse effects and steady-state plasma concentrations of amitriptyline were higher in non-individualised group during the whole treatment course. The results of our study suggest that the multiple points method is the most precise, but tedious and not practical. The modified Bayesian method with correction based on first measured plasma steady-state concentration of amitriptyline offers similar therapeutic outcome and adverse effects score combined with low cost and being easy-to-use.