The purpose of this research was to use Rasch measurement to study the psychometric properties of data obtained from a newly developed Diabetes Questionnaire designed to measure diabetes knowledge, attitudes, and self-care. Specifically, a methodology using principles of Rasch measurement for investigating the cross-form equivalence of English and Spanish versions of the Diabetes Questionnaire was employed. A total of fifty diabetes patients responded to the questionnaire, with 26 participants completing the English version. Analyses detected problems with the attitude items. We attributed the scaling problems to the use of negatively worded items with participants having generally low educational backgrounds. Analysis of the knowledge and self-care items yielded unidimensional variables with clinically meaningful item hierarchies that may have relevance to treatment protocols. Furthermore, the knowledge and the self-care items from the two versions of the Diabetes Questionnaire met our criteria for establishing cross-form equivalence and thus allow quantitative comparisons of person measures across versions. Limitations of the study and suggested refinements of the Diabetes Questionnaire are discussed.