A Norwegian version of this article has formerly been published in this journal:Walgermo, B. R., Uppstad, P. H., Lundetræ, K., Tønnessen, F. E., & Solheim, O. J. (2018). Kartleggingsprøver i lesing-tid for nytenking?. Acta Didactica Norge, 12(4), art. 7. https://doi.org/10.5617/adno.6499. Abstract At present, the national Norwegian quality-assessment system for basic education (NKVS) includes a range of reading tests with different purposes and designs, to be administered at various points during a 13-year period of schooling. In this article, we trace the historical background to the introduction of screening tests of reading in early education and identify a number of circumstances that call for a rethink of the overall test concept. Using data from a longitudinal study, we show that scores on a brief task administered towards the end of grade 1 predict reading-comprehension difficulties in grade 3. Taking that finding as our starting point, we discuss an idea for a possible new test concept that might have the potential to (a) improve the match between reading theory and reading-skill measurements, (b) enable longitudinal prediction, and (c) take less time to administer and be more useful in an educational context. The central elements of this new test concept are a short initial test meeting the primary objective of identifying students at risk of developing reading difficulties, and a follow-up explorative part to be carried out one-on-one, which will provide the teacher with information about the nature of each student’s difficulties. Keywords: reading assessments, screening for reading difficulties, longitudinal prediction, early efforts, dyslexia, reading and writing difficulties, special-needs education