Over two decades have passed since the federal policy on environmental justice (EO 12898) was issued. However, empirical evidence indicates that injustice persists and that US states vary in their adoption of the terms of the environmental justice (EJ) policy. Moreover, studies of the explanations for the variation in states’ adoption of EJ policy are rare and have yielded puzzling findings - e.g., environmental interest groups are not associated with states’ EJ policy adoption, or the severity of problems is associated inversely with their adoption. We examined the progress and variation in states’ EJ policy adoption as of 2005 using fuzzy-set qualitative comparative analysis. Our analysis showed first that a strong environmental interest group presence, combined with high racial diversity and low problem severity, is sufficient for a high level of EJ policy adoption, especially in Western states. Second, when environmental interest group presence is weak, if it is combined, again, with high racial diversity and the presence of a more liberal state government, a high level of EJ policy adoption also occurs. This is observed in the East coast, Midwestern, and Southern regions of the USA. Environmental politics and policy research can benefit from a configurational approach, especially when there is no guiding theory on the conjunctional effects of key factors.