IEA-studies on reading literacy of 9-10-year-old students provide an extensive source for trend analyses over a period of 35 years. A precondition for such analyses is to measure the students’ abilities on the same metric. Since the tests in the studies from 1970, 1991, 2001 and 2006 are not identical, this condition is not fulfilled from the outset. However, the IRT-technique provides an approach that allows to link tests and to establish a comparable measure on the same scale and over time. Such an analysis is based on the prerequisite that at least some of the items are equal in the different test forms and that they have similar psychometric properties over time. Bridges of items are available between all these studies by the links in the international studies, RL1991 and RL2001 (Martin, Mullis, Gonzalez, & Kennedy, 2003), PIRLS2001 and PIRLS2006, (Mullis, Martin, Kennedy, & Foy, 2007), and by Swedish extensions of the international designs which provide further bridges between the tests from 1970s reading comprehension and subsequent studies (Rosén, 2006; Taube, 1993). The research question addressed in this study concerns the reliability of these bridge items, i.e. to what degree they are sufficient to form a common IRT-scale.