The technique of multivariate analysis is particularly suited to educational needs assessment research because it allows for the summarization of data across any number of learners or components of educational need to produce a single numerical index of need for each skill examined. In the needs assessment process, educational or training need is assumed to have three underlying dimensions: competence of the individual at a task or skill, relevance of the task or skill to the individual, and the individual's desire to further his or her learning of the task or skill. The use of multivariate analysis as a means for assessing educational need is superior to previous methods in that these earlier models could accommodate data for only two dimensions. In addition, the multivariate analysis method manifests an increased sensitivity to changes in respondent distribution in the index of educational need. Finally, the model can accommodate different numbers and relative emphases of dimensions according to user-defined models and its method of computation is a relatively simple one. (This paper includes a discussion of the multivariate assessment technique as a model for assessing educational and training need, discussions of bivariate and multivariate cases of analysis, a computational example, a list of references, and six figures and one table illustrating various phases of the model and its standard error weights.) (MN)