Introduction. Bioequivalence (BE) studies are a necessary element of generic drug development. The primary BE criterion is the 90% confidence interval (CI) for the ratio of geometric mean pharmacokinetic (PK) parameters of the test and reference druds. Variability in the CIs calculation methods may lead to different results, the adoption of a false conclusion about the equivalence/non-equivalence of the compared drugs. The purpose of the work was to compare the algorithms for calculating the confidence intervals of pharmacokinetic parameters for assessing the bioequivalence of generic drugs. Materials and methods. Pharmacokinetic (pharmacodynamic) parameters obtained from studies performed in the organization in 2019-2024 were used. For parameters: maximum concentration or activity (Amax or Cmax), area by pharmacokinetic curve (AUC0-t from time zero to time of last blood collection and AUC0-∞ from time zero to infinity), Cmax/AUC0-t ratios (recommended only for veterinary products) CI were calculated with several algorithms formulated on the basis of regulatory documents and literature data. For statistical processing, Microsoft Excel 2007 (USA) and Statistica 10.0 (StatSoft, USA) were used. Results and discussion. In studies of generic drug for humans and animals with different routes of administration (oral, intramuscular, intravenous) using quantitative analysis of the active substances of the compared drugs by chromatographic methods, the reviewed algorithms made it possible to obtain similar CI values and uniform conclusion about the correspondence of the drugs to the hypothesis of their BE. A similar result was obtained for low molecular weight heparins drugs for two routes of administration (subcutaneous and intravenous) and two pharmacodynamic parameters (anti-Xa and tissue factor pathway inhibitor content), in contrast to the third recommended parameter - anti-IIa activity. Conclusion. Comparison of possible methods of CI calculating for BE studies showed that the algorithm based on the use of logarithmic transformation of PK parameters, estimation of the total error of the most significant factors (individual, drug, sequence, period) and their inverse transformation is the most relevant to modern regulatory documents and correct models of statistical data processing for such studies. It provides the narrowest range of CIs, presumably reducing potential risks to patients.