Determination of the heat transfer coefficient (HTC) distribution is important during the design and operation of many devices in microelectronics, construction, the car industry, drilling, the power industry and research on nuclear fusion. The first part of the manuscript shows works describing how a change in the coefficient affects the operation of devices. Next, various methods of determining the coefficient are presented. The most common method to determine the HTC is the use of Newton’s law of cooling. If this method cannot be applied directly, there are other methods that can be found in the open literature. They use analytical formulations, the lumped thermal capacity assumption, the 1D unsteady heat conduction equation for a semi-infinite wall, the fin model, energy conservation and the analogy between heat and mass transfer. The HTC distribution can also be calculated by means of computational fluid dynamics (CFD) modelling if all boundary conditions with fluid and solid properties are known. Often, the surface on which the HTC is to be determined is not accessible for any measuring sensors, or their installation might disturb the analysed phenomenon. It also happens that calculations using direct or CFD methods cannot be performed due to the lack of required boundary conditions or sufficiently proven models to analyse the considered physical phenomena. Too long a calculation time needed by CFD tools may also be problematic if the method should be used in the online mode. One way to solve the above problem is to assume an unknown boundary condition and include additional information from the sensors located at a certain distance from the investigated surface. The problem defined in this way can be solved by inverse methods. The aim of the paper is to show the current state of knowledge regarding the importance of the heat transfer coefficient and the variety of methods that can be used for its determination.