The Dutch government aims at reducing the percentage of low literates in the Netherlands (14.3%). In this study, we analyzed the reading component of the PISA 2009 data on the basis of three research questions: (1) Which reading items are relatively difficult for Dutch students in comparison with students from Flanders and Germany and which reading items are relatively difficult for low literates in comparison with non-low literates; (2) we compared the diversity and type of reading materials which low literate and non-low literate students read and how this affects their reading literacy; (3) we investigated differences between low literate and non-low literate students in the perceived usefulness of reading strategies. In order to answer the first research question, we calculated deviant items by subtracting the average difference in p-values between the Netherlands and the average for OECD-countries and dividing this by the standard deviation of all differences. For the relatively difficult items determined in this manner we investigated whether these were based on continuous or non-continuous texts and which reading competences they required. For answering the two other research questions we used regression analysis and chi-square tests.