This paper accepts that the OECD's PISA has become influential in policy terms globally, but analyses the ways that the main PISA and PISA for Schools tests are positioned differently in Australia and the USA because of contrasting educational federalisms in the two nations. Our argument is that while PISA is undoubtedly influential, its effects are nonetheless mediated by the political structures--here, the different models of federalism--present within different nations, which in turn leads to quite distinct "PISA effects". For instance, Australia oversamples on main PISA to make its data available for national and state-level policymaking, whereas the USA, with its focus on local governance in schooling, does not oversample, meaning that main PISA does not have a comparable policy salience as in Australia. Conversely, the newer PISA for Schools test originated in the USA with pressure from educators and philanthropic interests and has been implemented in a good number of schools, but it has not been taken up in the same way in Australia. Our analyses show how these differences reflect the idiosyncratic workings of federalism in the two nations, in which the federal government has a stronger policy and funding role in Australia than has hitherto been the case for the federal government in the USA.