<p>Mathematics achievement difficulties contribute significantly to student attrition from postsecondary STEM programs. Many programs offer diagnostic tests embedded in preparatory or introductory mathematics courses to facilitate early identification and remediation of gaps in foundational knowledge. However, the utility of these tests depends on students’ test engagement and motivation—factors that can vary both among individuals and across STEM fields, depending on the perceived relevance of mathematics in each field. Guided by situated expectancy-value theory, this study examined interindividual and context-dependent differences in STEM students’ test-specific motivation, engagement, and performance on a mathematics prerequisites test (<InlineEquation ID="IEq1"> <EquationSource Format="MATHML"><math> <mi>N</mi> <mo>=</mo> <mn>2384</mn> </math></EquationSource> <EquationSource Format="TEX">$N = 2384$</EquationSource> </InlineEquation>; 31 courses). Multilevel analyses revealed substantial between-course variation in test-specific achievement and motivation linked to course-level differences in prior mathematics achievement and motivation. The perceived relevance of the tested mathematics content also differed across courses and programs. Within courses, students who perceived the tested mathematics content as more relevant to their study program than their peers performed better and reported a higher content-specific self-concept of ability and intrinsic value, even after controlling for prior achievement and general mathematics motivations. Beyond this individual effect, students in courses with a higher perceived relevance reported higher test motivation and showed higher test engagement and performance (indicating a positive context effect of relevance). These findings emphasize the importance of aligning diagnostic tests with field-specific perceptions of mathematical relevance to enhance student motivation, engagement, and ultimately the utility of diagnostic testing across STEM disciplines.</p>