Programmers using imperative languages have a number of well-established debugging tools available to them; functional programmers have few, if any, tools available. Many of the tools and techniques developed for debugging functional programs are based on those for imperative programming and lack a theoretical basis relevant to functional programming. In addition, the techniques used are typically very time-consuming. A theoretical foundation on which to base the study of errors and debugging in functional programming is presented in this paper. Using this theoretical foundation, a set of program transformation schemes has been developed which facilitate the location of the type of error which results in an evaluation-time error message and the termination of evaluation. A brief description of the practical experience obtained using the tool is also presented.