This study investigated changes in reported test-taking motivation from a low-stakes to a high-stakes test and if there are differences in reported test-taking motivation between school classes. A questionnaire including scales assessing reported effort, expectancies, perceived importance, interest, and test anxiety was administered to a sample of pupils (n = 375) in 9th grade in direct connection to a national test field trial and then again to the same sample in connection to the regular national test in science. Two-level second-order latent growth modelling was used to analyse data. In summary, the results show a significant increase in reported test-taking motivation from the field trial to the regular test and a significant variability in test-taking motivation between classes.