During the past decade, it has become fashionable to include an evaluation component with each new bilingual education program. The proliferation of empirical evaluation studies seems, however, not yet to have shed much light on very basic issues, such as the relationship between language of instruction and cognitive growth, academic achievement or the development of reading, writing and speaking skills. Furthermore, it has become increasingly obvious that political and social pressures are usually more important factors in producing changes in educational policy than the results of empirical research. In this paper, examples from programs in Canada, Haiti, Nigeria and the Philippines are drawn upon to examine issues such as the following: (1) the reasons why evaluations are typically undertaken; (2) the groups or individuals from whom pressure for evaluation comes; (3) the types of evaluation strategies that have been used successfully in various places; (4) the major components of such evaluations; (5) the ways in which the results of evaluations filter down to parents, educators, administrators and political decision makers; and (6) the type of technical skills an evaluator should possess. (Author/CLK)