The 'national curriculum' announced by the 1988 Education Act specifies that in the later years of secondary schooling all pupils must take courses in maths, English, science, religious education, history, geography, technology, music, art, physical education and a second language. One of the consequences of this new curriculum seems likely to be a restriction of the amount of choice that pupils (and indirectly their parents) have in selecting fourth and fifth year courses in secondary schools. Thirty years ago, pupils had very little choice in this area. But, with the growth of large comprehensive schools and the expansion and diversification of national examinations, option choice emerged as a key point in the secondary school careers of pupils. At face value, the new Act heralds a return to the earlier situation. However, from another point of view the Act may represent little change. Some studies of the option process have suggested that in practice pupils were given very limited choice of courses, and that those in lower streams or bands had virtually no choice at all. ' If what Woods and Ball claim is true, the National Curriculum may have little practical effect on the range of course choice available to pupils in secondary schools. Before reaching this conclusion, though, we must ask ourselves how sound a basis these studies of option choice provide for drawing such a conclusion. Do they give us an accurate account of the operation of option choice systems? In this paper I shall examine the claims that Woods and Ball make in their studies of option choice, and the evidence they offer in support of these claims. An underlying concern is to assess the contribution that ethnographic studies can make to the understanding of public policy issues. I would not want to suggest that any definitive conclusion about this larger question should be made on the basis of an examination of just two studies in a particular area. However, examination of these studies will, I hope, highlight some of the strengths, weaknesses and problems surrounding the contribution of ethnographic research to policy debates.