Five elements necessary for the rigorous evaluation of bilingual programs are discussed: (1) careful collection of meaningful baseline data from selected subjects; (2) the identification and development of instruments to measure key variables, such as context and student characteristics; (3) the identification of treatment characteristics, such as curriculum design and materials, or the method of language usage; (4) the establishment of longitudinality; and (5) the interpretation of results in implementable terms that are meaningful to teachers, policy makers, and researchers. Other considerations discussed are the validity of available tests, data collection and management, and statistical analysis of the data. (BW)