The goal of this study is to find out whether it is possible to explain the cross national differences in students' literacy in science, with the available PISA(Programme for International Student Assessment) data. This goal was approached by analyzing the possible relationships of the context indicators on different levels(student and school level), with scientific literacy in an international comparative context. For this, the PISA 2006 data from three relevant reference countries(Korea, Finland, and Hong Kong-China) were explored. Multi-level statistical models were used for the estimation of the parameters by means of the software HLM, version 6.0(Raudenbush, et al., 2004). Both similarities and differences were found with regard to predictors at student and school level of scientific literacy. It turned out that school background variables are closely related to scientific literacy both at school levels for Korea and Hong Kong, but not for Finland. A limited number of 'changeable' school variables seem to be related to achievement in all three countries. On the other hand, students' motivation or self-related belief factor is an important 'changeable' influencing factor in all three countries. In summary, these results partially explained cross-national differences in relations between potentially effective factors on scientific literacy. If the conceptual foundation of the multi-level models for scientific literacy could be improved, it would be provided better opportunities to explain differences between countries in achievement. However, the search for reasons why one country is outperformed by another country still remains very complicated because of the many factors involved and the social and cultural differences between countries.