The amount of water in pharmaceutical dosage forms (PDF) is determined by different experimental techniques. One of them is water activity (aw) measurement, which gives information about the amount of free water in PDF. Free water is available for chemical changes, physical interactions and for microbiological growth. By definition, aw is the ratio between the vapor pressure of water above the sample and the vapor pressure of pure water at the same temperature. Different types of devices are used to measure aw, which operate on different measurement principles. In the master's thesis we examined how different factors influence the determination of aw in PDF. For this purpose, we first developed and validated aw measurement methods for four selected PDF (powder mixture, hard gelatin capsules, HPMC capsules, tablets). Then we compared the obtained sorption isotherms of PDF to the theoretical calculations from the sorption isotherms of individual excipients and mass fractions. We found very good agreement between experimental data and theoretical prediction for all four tested PDF. In this section, we also examined the influence of the technological process on the sorption isotherm. By comparing the sorption isotherms of the powder mixture and tablets containing the same qualitative and quantitative composition, we showed that the tableting does not affect the sorption isotherm. Next, by examining the sample of gelatin capsules, we confirmed the theoretical equality of aw for all the components of the PDF in equilibrium. We also evaluated the influence of volatile organic solvents (ethanol, isopropanol, acetone) on the determination of aw, which were added to variously moistened samples. We found that the residual solvents at the level of the permissible limit in accordance with the ICH Q3C guideline affect the determination of the aw. Finally, we checked the influence of the relative humidity of the room in which we performed the measurements of aw. We concluded that it makes sense to perform measurements at a relative humidity that is as similar as possible to the aw of the measured sample. Measurements were performed on two devices (AquaLab, LabMaster), which measure aw by different principles. Based on the obtained results, we concluded that both devices give comparable results. Additionally, we also showed for both devices that residual solvents in the final products interfere with the measurement of aw.