The use of lithium in medicine began in the mid-19th century, when the solubility of uric acid salts in the solution of lithium carbonate was demonstrated in vitro, which meant that lithium could be used in the treatment of gout. The use of lithium in psychiatry starts in 1949, when the effect of lithium in the treatment of mania was demonstrated. The mechanism of action of lithium is not yet completely understood. In parallel with the use of lithium in the treatment of psychiatric disorders, different methods for the determination of lithium in human samples have been developed. The first aim of the study was to determine the amount of lithium in serum samples using electrochemical methods, flame photometry and spectrophotometry and use the results to compare the results obtained using these methods. Because of the possibility of determining lithium in various media, the second aim of the study was to evaluate the clinical value of determining the concentration of lithium in erythrocytes in relation to the concentration of lithium in plasma. The third aim was to investigate the extent to which a therapeutic dose of lithium correlates with the measured concentrations of lithium in serum, plasma and erythrocytes. It was concluded that statistically there was no significant difference between the three test laboratory methods (P = 0.507). Investigating the correlation between the concentration of lithium in various media measured by different methods and the daily therapeutic dose of lithium, it was concluded that statistically a significant correlation was found only in serum lithium concentrations measured with electrochemical method (P = 0.009; r = 0.47). There was statistically a significant moderate correlation between the concentration of lithium in plasma and erythrocytes (P = 0.002; r = - 0.54), and the lithium concentration erythrocytes are higher than lithium concentrations in plasma (P = 0,043). The range of the ratio of the concentration of lithium in erythrocytes and plasma is wide (13.25 to 111.15), and is not in correlation with the therapeutic daily dose and therefore is not a better indicator in the control of the treatment.