Despite a large body of literature on self-regulated learning and metacognitive strategy use, relatively few empirical studies were situated in authentic assessment settings to explore the types of strategies employed by L2 learners, especially L2 writers, and how such strategy use may affect their assessment performance. Using a post-test metacognitive writing strategies questionnaire, this study therefore examined the type and effect of metacognitive writing strategy use among about 200 EFL students immediately after completing the writing assessment on a university placement test. Factor analysis of empirical data pointed to five main types of strategies employed by EFL writers when composing under a timed testing setting. Correlation and regression analysis, however, revealed limited and mixed effect of such strategy use on writing assessment performance. Results from this study call for a more nuanced and sophisticated treatment and measurement of metacognitive strategy use in future research, particularly in instrument design and threshold effect testing. Such new evidence can inform writing pedagogy, especially in the face of a recent call for integrating metacognitive strategy instruction into L2 writing classrooms.