Functional data have been both the least used and the least understood class of data in systematics. Compared to the use of morphological features from DNA sequences to gross structural characters, patterns of distribution, and even ecological and behavioral attributes of organisms, functional characters have not been generally thought of as useful for resolving systematic and historical questions. The current status of functional data in systematics is due to three primary factors, both historical and practical. First, current research in systematics developed primarily out of the nineteenth century morphological tradition while the analysis of organismal function was centered in physiological and experimental embryological research. At the turn of the century these two research traditions, initially complementary, diverged. While several authors have debated the severity of the divergence in research between morphologists (interested in the comparative analysis of structure, its development, and phylogeny) and experimental biologists (interested in function, physiology, or uses of structures) (5, 6, 102), there is little doubt that these two research areas diverged in the early part of this century and have remained largely separate. Second, functional data are hard to obtain on a diversity of taxa: Gathering a range of experimental data on even one small clade is time consuming, and conducting manipulative experiments to understand causal relationships only adds to the difficulty of comparative functional analyses. Third, many so-called "functional" analyses and discussions of the import of function for systematics are in fact purely morphological. The adjective "functional" has acquired a cachet: It sounds quantitative, technical, and sophisticated. As a consequence, it is frequently used by papers in which there is no resemblance of a true functional analysis, where organismal function is directly measured and compared with measured function in other clades. The valid heuristic use of functional ideas should not be a substitute for direct measurement. A key aim of this paper is to suggest that functional data are in fact critical to understanding five important issues in systematics. Despite difficulties in gathering functional data, the development of even a small number of well-understood case studies would greatly enhance our appreciation of (a; historical patterns to functional transformation, (b) the use of functional characters to define monophyletic clades. (c) the causal basis of character distributions on cladograms, (d) the extent to which changes in structural, functional, and behavioral characters are historically congruent, and the extent of evolutionary conservatism at any particular level of organismal design, and (e) general patterns and principles in the evolution of form and function. [ABSTRACT FROM AUTHOR]