This report arises from an earlier one produced for OECD on the Use of Longitudinal Data in the study of Social Exclusion (Bynner, (1996). That focused on the availability of large scale and long term longitudinal data as a research resource in the study of social exclusion. This report is concerned with findings from studies using longitudinal data that contribute to our understanding of the topic. I was asked to do this through the linkages between disability and disadvantage to education and employment and other "socially excluding externalities". The field is potentially a huge one so I have been selective in restricting the report to studies that are complementary in adding significantly to knowledge of social exclusion - especially as manifested through backwardness in education and its counterpart - behaviour problems and criminality in adolescence. This has merited a fairly extensive use of research - much of it involving different types of longitudinal data - to examine the concept of social exclusion and how it relates to ideas of risk and protection. The first part of the report is devoted to this question. The second part focuses on the idea of childhood risk and protection. The next two sections examine in more detail findings from longitudinal research, first on risk and protective factors and second on life course patterns leading to social exclusion in adulthood. The final section draws some general conclusions about the social exclusion process and considers some implications for policy.