It has often been observed that there is a significant gap between the aspirations for the teaching of English as expressed in Japanese Ministry of Education curriculum guidelines and the reality of classroom practice. Using a combination of questionnaires, semi-structured interviews, classroom observations and curriculum and textbook analysis, this thesis seeks to explore that gap. While many of the reasons identified in the past (community expectations, the negative impact of entrance examinations, and limited time for lesson preparation) did emerge as being of significance, a wide range of other issues also emerged which, taken together, suggest that the problems faced by teachers of English in Japanese schools are more complex and multi-faceted than the literature appears to suggest. Analysis of the Japanese curriculum for English indicates that while it is clearly influenced by developments in the areas of communicative competence and communicative language teaching and by research in the area of discourse analysis, it includes many features that are reminiscent of a considerably more traditional approach. It appears to proscribe in some places what it recommends in others, provides little guidance on critical aspects of curriculum implementation, and was judged by some of the teachers involved in the study to take little account of the context in which Japanese teachers work. So far as language teacher training is concerned, there appear sometimes to be very significant gaps in what is provided, with the courses offered often being taught by academics who may, in some cases, have themselves had little training in language teaching and may also lack understanding and experience of teaching in schools. Widely used textbooks, all approved by the Ministry of Education and written by teams dominated by university-based academics, appear to be largely absent of any genuine communicative orientation. Add to this the fact that changes in expectations relating to teacher behaviour have not been accompanied by any concerted effort to change community attitudes or outdated testing and assessment procedures, and it should come as no surprise to find that the language lessons observed were teacher-dominated, with the teachers talking, mostly in Japanese, for most of the time, and with considerable reliance on translation, repetition and rote learning. Although it seems to be widely believed that grammar translation is the dominant approach, these lessons exhibited a curious mixture of aspects of grammar translation and aspects of audio-lingualism (with its behaviourist and structural underpinnings). It has often been noted that teachers in Japanese secondary schools are generally heavily burdened with non-teaching responsibilities. However, the constant teacher-focus and the ongoing struggle to communicate with students that characterized these lessons would appear to do little to ease the burden on teachers. In spite of all of the problems they face, many of the teachers involved in this study appreciate the value of making language lessons interesting and indicated that they are ready and eager for change. If change is to happen, the Ministry of Education will need to acknowledge that teachers cannot be expected to take full responsibility for it. A strategic approach to what is clearly a systemic problem is required.