Brachytherapy is a specific modality of radiation therapy used for treating oncology patients, in which encapsulated radioactive source of ionizing radiation is inserted into or near the tumor volume in order to kill tumor cells. In this Diploma thesis the emphasis is on high dose rate brachytherapy performed by the method of remote afterloading of the source, iridium; 1923^Ir. The dosimetry accuracy of high dose rate brachytherapy depends significantly on appropriate quality assurance procedures, among which the primary procedure is to verify the magnitude of the reference air kerma rate (RAKR) given by the source manufacturer. RAKR is recommended quantity for describing strength of photon-emitting brachytherapy sources and an important input parameter for computer system which calculates and optimizes the absorbed dose distribution. In this thesis the large set of relative discrepancies between the RAKR 192^Ir, measured in the hospital and RAKR of the same source given by source manufacturer are retrospectively analyzed. Data were collected based on regular measurements with a well-type ionization chamber, conducted over a long period of time at the Clinic for Oncology and Nuclear Medicine, KBC Sestre milosrdnice in Zagreb. The results of RAKR measurements and comparison with the manufacturer determined RAKR value show good agreement and that the established measurement procedures are reliable - the mean relative discrepancy is -0.09% with a standard deviation of 0.95%, and all discrepancies are within the recommended range ± 3%. The discrepancies were further divided into 3 groups based on the change of source construction and the afterloading unit used, and the change of the RAKR measurement protocol; from one source insertion into the ionization chamber per measurement, to three source insertions per measurement. Analysis of variance (ANOVA) showed that there is no difference between the mean discrepancies of the measured RAKR from the manufacturer's specification, if the discrepancies are classified into three groups. The estimated combined expanded uncertainty of the measured RAKR has value 2.88% with a coverage factor k=2.