We test the ‘law matters’ and ‘legal origin’claims using a newly created panel dataset measuring legal change over time in a sample of developed and developing countries. Our dataset improves on previous ones by avoiding country-specific variables in favour of functional and generic descriptors, by taking into account a wider range of legal data, and by considering the effects of weighting variables in different ways, thereby ensuring greater consistency of coding. Our analysis shows that legal origin explains part of the pattern of change in the adoption of shareholder protection measures over the period from the mid-1990s to the present day: in both developed and developing countries, common law systems were more protective of shareholder interests than civil law ones. We explain this the result on the basis of the head start common law systems had in adjusting to an emerging ‘global’ standard based mainly on Anglo-American practice. Our analysis also shows, however, that civil law origin was not much of an obstacle to convergence around this model, since civilian systems were catching up with their counterparts in the common law. We then investigate whether there was a link in this period between increased shareholder protection and stock market development, using a number of measures such as stock market capitalisation, the value of stocktrading and the number of listed firms, after controlling for legal origin, the state of economic development of particular countries, and their position on the World Bank rule of law index. We find no evidence of a long-run impact of legal change on stock market development. This finding is incompatible with the claim that legal origin affects the efficiency of legal rules and ultimately economic development. Possible explanations for our result are that laws have been overly protective of shareholders; transplanted laws have not worked as ex- pected; and, more generally, the exogenous legal origin effect is not as strong as widely supposed.