International assessments are often developed in one country and applied in other countries. Assessing the measurement invariance across countries is an important step in determining if valid conclusions can be drawn in these other countries and if valid comparisons can be made across countries under these circumstances. This study investigated measurement invariance, across two countries, of selected questions from the Programme for International Student Assessment 2009 student questionnaire. Turkey and United States were compared by using multiple group confirmatory factor analysis for scores on polytomous items. Specifically, the questions concerned enjoyment of reading, learning strategies, teacher student relations, and disciplinary climate, teachers\U+2019\ stimulation of reading engagement, and teachers'use of strategies. The results indicated that based on the chi-square goodness of fit test and root mean squared error of approximation, none of the models proposed in OECD (2012) adequately fit the data and all questions exhibited differential item functioning (DIF). However, based on the comparative fit index and the Tucker-Lewis index, models for questions, other than except learning strategies, fit the data. For these questions, results supported the fit of the strict factorial invariance model and indicated that the questions functioned similarly for students in Turkey and the United states. Thus the evidence for DIF was ambiguous except in regard to learning strategies. Under the assumption that the questions functioned similarly in Turkey and the United States, factor mean were compared across the two countries on questions other than learning strategies. According to the comparison, Turkish and American students significantly differed on all questions. The difference was large for enjoyment of reading, moderate for teachers' ruse of strategies, and small for the remaining questions.