Driven by a desire to improve academic outcomes and transform "failing" schools, governments around the world have often turned to the development of new forms of state-funded school. This paper looks at three such instances of the introduction of new forms of schooling, within three urban localities (academy schools in London; charter schools and small schools of choice in New York City; and Schools of Tomorrow in Rio de Janeiro). It considers the extent to which these types of school did improve academic outcomes for their students and draws comparisons across each case study in order to understand their similarities and differences. It concludes that although the quasi-marketisation of school systems through the introduction of new (often private) providers might improve outcomes, this is not the only means by which improvement can be attained; and that instead the introduction of new forms of school may be successful because this enables certain other changes to happen. It highlights the limited nature of impact evidence available in all instances, which restricts our ability to properly evaluate the effect of new school types on outcomes.