Techniques for solving inverse problems as well as their applications are currently rapidly developing in all the different domains of physical sciences and particularly in Heat Transfer. These techniques are usually developped by applied mathematicians, statisticians and signal processing specialists. Experimentalists desiring to go beyond traditional data processing techniques for estimating the parameters of a model with the maximum accuracy feel often ill-prepared in front of inverse techniques. This paper is devoted to the presentation of the common methodology that can be used by a heat transfer specialist in order to solve either a parameter estimation problem (thermal characterization of a material for example) or a function estimation problem (estimation of a transient heat flux at the front face of a wall for example), starting from transient temperature measurements inside or on the external boundaries of the physical system. Once this common methodology shared by specialists of measurement inversion techniques, modeling techniques and experimental techniques, a very wide variety of tools (analysis of the sensitivity matrix, variance-covariance matrix of estimates, regularization and Bayesian techniques) are now available in order to avoid biases at different levels of this kind of involved task.