General practitioner fundholding is often represented as one of the more successful elements of the 1989/90 Conservative reforms of the UK National Health Service (NHS). Successive annual ‘waves’ of fundholding practices were approved from 1990 through to 1997 and, over time, the initiative came to involve some 50% of UK generalpractitioners. Fundholding is known to have had a strong regional geography that changed with evolving fundholding eligibility criteria. Further, there have been persistent allegations that fundholding tended to occur disproportionately in areas of higher social status. Past studies of fundholding have tended to consider single waves or the overall impact of the initiative rather than its development over time. They have also tended to work at a single geographic scale or through single-region case studies when exploring the statistical regularities underlying the uptake of fundholding. Using multilevel analysis, this paper seeks to enhance understanding of fundholding through an examination of the interaction of district health authority and practice characteristics across all implemented waves for all general medical practices in England and Wales. We conclude that wave mattered on a national scale, that deprivation was relatively unimportant and that there were certain types of area that exhibited persistent but unexpected high uptake.