Since improvements in cancer screening, diagnosis, and therapeutics, cancer disparities have existed. Marginalized populations (e.g., racial and ethnic minorities, sexual and gender minorities, lower-income individuals, those living in rural areas, and persons living with disabilities) have worse cancer-related outcomes. Early detection of cancer substantially improves outcomes, yet uptake of recommended cancer screenings varies widely. Multi-cancer early detection (MCED) tests use biomarkers in the blood to detect two or more cancers in a single assay. These assays show potential for population screening for some cancers—including those disproportionally affecting marginalized communities. MCEDs may also reduce access barriers to early detection, a primary factor in cancer-related outcome disparities. However, for the promise of MCEDs to be realized, during their development and testing, we are obligated to be cautious to design them in a way that reduces the myriad of structural, systematic, and personal barriers contributing to disparities. Further, they must not create new barriers. Population studies and clinical trials should include diverse populations, and tests must work equally well in all populations. The tests must be affordable. It is critical that we establish trust within marginalized communities, the healthcare system, and the MCED tests themselves. Tests should be expected to have high specificity, as a positive MCED finding will trigger additional, oftentimes invasive and expensive, imaging or other diagnosis tests and/or biopsies. Finally, there should be a way to help all individuals with a positive test to navigate the system for follow-up diagnostics and treatment, if warranted, that is accessible to all.