<p>While research on writing has long emphasized motivation as a key predictor of performance, scholars are increasingly exploring whether engagement, as a more immediate and instructionally actionable construct, might offer a more useful lens for understanding how students succeed. The present study compared the relative predictive power of writing motivation and writing engagement in undergraduate composition. Students across multiple U.S. institutions completed the Writing Engagement Scale–College (WES-C; <i>n</i> = 530), and a subset also completed the Writing Motivation Scale (WMS; <i>n</i> = 307). Writing performance was assessed via rubric-based ratings of course-embedded essays for a complete-case subsample from one institution (<i>n</i> = 96). Correlational and regression analyses compared the contribution of motivation and engagement to text quality, and a bootstrapped mediation model tested whether engagement mediated the motivation–performance relationship. In the performance subsample, engagement explained substantially more variance in writing quality than motivation, and when both predictors were entered simultaneously, engagement remained significant while motivation did not. Mediation results further indicated a significant indirect effect of motivation on writing quality via engagement. While motivation remains theoretically important, its unique predictive contribution diminished once engagement was taken into account. These results clarify the distinction between the constructs and suggest that task-level writing engagement provides a more proximal and instructionally interpretable lens for understanding writing performance.</p>