For decades the world society perspective has influenced comparative research on a broad range of issues across the social sciences. The perspective emerged to make sense of an empirical puzzle: why did nation-state after nation-state expand mass schooling after World War II? The perspective evolved to address broader issues such as the authority of science and its influence on the environmental movement, the expansion of the scope of citizenship and its impact on women's rights and, more recently, the rise of an international human rights regime on the one hand and the celebration of universities of excellence on the other hand. The world society perspective has motivated research that examines worldwide and regional trends and generates hypotheses to explain these cross-national developments and variations. This paper first clarifies some of the world society perspective's key assumptions and core arguments as applied to comparative education. Next, the paper situates the world society perspective within a broader neo-institutional theoretical framework. Neo-institutionalists assume that the actors are highly embedded in their larger environment. Following this assumption, we distinguish between levels of analysis (macro or societal, meso or organisational, and micro or individual levels) and the extent to which the relevant actors are conceptualised as embedded in environments. The paper then turns to consider the character of the influential environment and to discuss the role of models of reality as rationalising and legitimating myths. From a neo-institutional perspective, the actors are often imagined as enacting scripts that make sense given the triumph of some models of reality. The fourth section of this paper focuses on the concepts of institutional isomorphism and loose coupling and their uses in comparative research guided by the world society perspective. Lastly, the paper clarifies the kinds of research strategies associated with the world society perspective, from trend identification to more explicit efforts to model the diffusion of discourse, policies, structures, and practices. These strategies have varied over time. A common thread though is to empirically ascertain as to whether there is evidence supporting the dynamics emphasised within the world society perspective, net of other influences that impinge on the outcomes of interest.