This paper presents a method for estimating the parameters appearing in heat transfer models. The method involves minimizing a function that is more general than that utilized in least squares. One of the unique features of the method is that it is sequential. Some of the advantages of this feature are that the basic equations are simple in that no matrix inverses are required, computer programs can be easily written to have an arbitrary number of parameters, and the sequential procedure provides information for development for better mathematical models for describing complex phenomena. Some examples are given.