Nowadays, in various industries, in particular in nuclear energy, to determine the fracture toughness, along with standard tests of compact specimens, which are quite expensive and complex, methods are developed to determine these characteristics by impact tests of Charpy specimens using different correlations between Charpy impact fracture energy (CVN) and critical stress intensity factor (J-integral). The paper analyzes correlation and analytical methods, the authors of which consider them universal for a certain class of steels. Correlation methods are divided into one-stage and two-stage. One-stage methods allow to obtain the value of the critical stress intensity factor by the known fracture energy. Two-stage methods in the first stage offer the calculation of the dynamic critical stress intensity factor, in the second the temperature shift and obtaining a static critical stress intensity factor. Analytical methods according to the іmpact fracture diagram of the specimen allow to construct a J-R curve and calculate the value of the J-integral. A series of fracture tests of CT specimens made of heat-resistant steel 22K was carried out, the reference temperature T0 was determined according to the single-temperature method of the ASTM-1921 standard and the Master curve was constructed. A series of standard Charpy specimens impact tests in the temperature range -50…+100°С was performed using an instrumented drop-weight impact testing machine equipped with a high-speed registration system. According to the results of Charpy specimens impact tests, the fracture toughness were determined using different methods. It is established that both analytical and correlation methods cannot be universal and can be used to determine the fracture toughness of 22K steel. Therefore, a new exponential correlation was proposed between the fracture energy of the Charpy specimens and the critical stress intensity factor for heat-resistant steel 22K.