This paper considers issues relating to the measurement of skill for national statistical purposes. It draws upon the work program and research underlying the revision of the national occupational classification for the United Kingdom (UK), SOC90 (Standard Occupational Classification introduced in 1990). The report's introduction states the intention to reflect upon the review-related research findings; detail the perceived inadequacies of SOC90; describe the problems associated with occupational definition in certain areas; show how the revised classification will affect the analysis of skill change; and cause experts to rethink the forecasts of occupational change. Section 2 presents an overview of the history of occupational classification in the UK. Section 3 describes the conceptual basis of the SOC. Section 4 details the perceived weaknesses in SOC90 and the constraints surrounding the development work undertaken to revise this classification. Section 5 outlines some key processes that were influential in bringing about a redefinition of occupations for statistical purposes. Section 6 discusses the resources that were used to investigate the processes of occupational change from a statistical and definitional perspective. Section 7 examines the revised classification in terms of its ability to distinguish and discriminate between occupations and the new analytical opportunities it will provide. Section 8 concludes that SOC2000 (published in spring 2000) makes better use of its conceptual base, solves problems inadequately dealt with earlier, and provides a better tool for job matching purposes than did SOC90. (YLB)