This paper presents the findings from a two-year research project which focused on coursework and coursework assessment in the GCSE. Case studies of the effects of coursework were made in six schools, across three different counties and two metropolitan districts. Coursework practices were found to differ, with variation in the following: timing of coursework during the two years, where it was being completed, the type of exercises that students were doing, the amount and quality of teacher input, the availability of resources, and the extent of parental and 'other' help. This was acting to decrease the examination's reliability as a testing device, while at the same time increasing validity. Maturation issues were found to be a problem in the making of reliable and valid assessments. Teachers were finding it difficult to reconcile contradictory demands -- the need to initiate a formative process of assessment and learning throughout the duration of the two-year course, and the requirement to undertake a summative process of assessment and reporting. Coursework has allowed parents to play a fuller and more direct role in the completion of coursework assignments, but parental interventions were found to be limited. But though one form of reliability is threatened by the introduction of coursework techniques, there are pedagogic, learning and motivational gains to be made. The GCSE was designed to integrate more productively coursework assessment techniques with programmes of study. But evidence from the case studies suggests that close integration of assessment task and learning programme was not being achieved. Finally, it was noted that some teachers were treating coursework in connected ways; whereas others, conscious of the need to assess in nationally equivalent environments, were formalizing the process, and as a result disconnecting assessment from learning and thereby limiting its notional ability to act formatively.