A therapeutic trial of antifungal therapy in humans requires not only well-defined objectives and a sufficient number of patients, but also standardized criteria for the diagnosis and evaluation of response, including the efficacy, failure, and toxicity of any new agent or treatment modality. Generally, the retention of a patient in a study should require, as a minimum, confirmation of the etiologic diagnosis by culture of the organism from the infected site. Two different systems for evaluation of therapy are proposed. The first is a set of definitions of response to therapy; these are adaptable for use in any type of prospective clinical trial and emphasize cure as the best possible outcome. The second system of evaluation employs scored criteria and thereby enables quantitative comparison of the status of a patient before therapy with his or her status at intervals during therapy, at the end of therapy, and at intervals after therapy. This system provides an objective means of measuring improvement in a given patient; thus it is most useful in those diseases in which significant improvement rather than cure is a more realistic goal of therapy, e.g., coccidioidomycosis.