We began this work with an intent to create an assessment strategy for community-engaged learning focused on sustainability. Through reviewing existing approaches, we discovered that the salient learning modality for such a purpose—transdisciplinarity—differs profoundly from traditional engineering education settings. Further, this form of community-engaged learning offers an authentic setting to develop many of the integrated student outcomes stated in the new engineering accreditation criteria. We contend that effective transdisciplinary learning is a means to the oft-stated goal of systemic transformation in engineering education, particularly for sustainability aims. However, this complex, dynamic systems view of engineering education represents a radical departure from education-as-usual and thus requires a similarly radical departure from research- and assessment- -as-usual. It reflects a shift in the unit of analysis: from a singular focus on student learning outcomes to a broader view that captures learning at the transdisciplinary system level. It also requires a shift from assessing/judging student learning based on learning outcomes alone to that of navigating the learning process. This paper opens a dialogue on a compelling question for engineering education: How do we design and navigate dynamic systems of learning that themselves create sustainable communities? We suggest an approach that includes monitoring system-level properties rather than a singular focus on the learning at the individual level. A review of literature reveals nascent alignment amongst researchers on core sustainability competencies needed for student learning, however, there is no consensus on systems-level learning approaches. Based on a literature review and lived experiences, we offer a unifying framework through a systems-level lens for essential properties of assessing community-engaged learning for sustainability: holistic, participatory and adaptive. We also offer reflections on qualities of transdisciplinary research approaches that are coherent with complex dynamic systems and suggest future directions.