An educational approach about literacy known as whole language has been popular and appealing to many educators in Taiwan. However, the fact that it is one of the most controversial issues in the history of literacy education is seldom reported. Through an extensive literature review, the present authors examine the theoretical bases and instructional issues of whole language, followed by a review of critiques from psychological and educational researchers. Most empirical studies do not support two important assumptions underlying the whole language approach: (a) the acquisition of written language is similar to that of spoken language, and (b) skilled readers rely on contextual information more than on the printed words. Based on meta-analysis studies examining effects of different reading instructions, as well as on the unsuccessful case of California education system which once embraced whole language, many researchers are skeptical, if not strongly opponent, to the claim of whole language effectiveness on literacy achievement. The present authors, however, hold a view that the effects of whole language might be a function of students' personal/family variables. Along with this reasoning, the authors also scrutinize the effects of whole language on disadvantaged, low-achieving children. It is concluded that disadvantaged and underachieving children may not benefited from the instruction of whole language approach.