In Sierra Leone, the number of primary schools is almost seven times more than junior-secondary schools (JSS). Living in a poor community has been shown to reduce children's school attendance because of lower access and poorer quality of education in these communities. However, it is unclear whether living in a poor community reduces attendance at primary school more than at JSS, given the wide disparities in school supply between the two educational tiers. Random effects logistic regression models are applied to the 2008 Sierra Leone Demographic and Health Survey for analyses. The results show that living in a poor community reduces primary school attendance but increases attendance at JSS. These results reflect post-war reconstruction in Sierra Leone where: (1) primary schools were rapidly reconstructed without inspection, leading to poor-quality schools; and (2) the redevelopment of JSS was concentrated in war-torn areas, which had minimal access to post-primary education in the pre-war era.