Coaching is known to improve student performance on tests with high personal relevance (“high-stakes tests”). We investigate whether the same holds for a test that has no personal relevance for the students taking it (“low-stakes test”). More specifically, we explore whether student performance on the reading and mathematics assessments of the OECD's Programme for International Student Assessment (PISA) can be fostered by coaching (and administering a pretest). Coaching and pretest effects were studied for each content domain separately in a pre-/posttest quasi-experimental design. To examine differential effects of academic tracks, samples were drawn from German Hauptschule and Gymnasium schools. Results show that only the combined effects of pretesting and coaching have substantial positive effects on student performance. Implications for the interpretation of large-scale assessment programs are discussed.