Formative assessments are primarily used as a tool to gauge learning throughout an anatomy course. They have also been demonstrated to improve student mastery and exam performance, although the precise nature of this relationship is poorly understood. In this study, it is hypothesized that formative assessment questions targeting higher cognitive levels, integrating topics from multiple lessons, and including visuospatial elements will increase student exam performance. Formative and summative questions provided to students during the Clinical Anatomy block at the University of Arizona College of Medicine--Phoenix between 2015 and 2018 were assessed for cognitive level, integration of targeted learning objectives, and presence or absence of visuospatial elements. These variables were entered into a hierarchical linear model along with demographic variables for each cohort to assess the relationships between these variables and cohort performance on exam questions. The best predictor of exam performance was the inclusion of constituent learning objectives within the formative assessment. Additionally, students performed better on exam questions with visuospatial elements when the targeted learning objectives were also associated with visuospatial elements on the formative assessment. Surprisingly, the cognitive level of formative questions and the integration of learning objectives within them were not correlated with student exam performance. This study demonstrates the importance of including a broad range of topics in formative assessments and highlights a potential benefit of adopting consistent question formats for formative assessments and exams.