The Centers for Medicare and Medicaid Services (CMS) coverage with evidence development (CED) program provides coverage for items and services not meeting Medicare's "reasonable and necessary" standard while requiring participation in clinical studies. As additional evidence is available, CMS may reconsider CED decisions. Of twenty-six items and services in the CED program since its 2005 inception, CMS has reconsidered coverage for ten (38 percent). Of 235 publications that CMS cited as evidence supporting reconsideration, 57 (24 percent) were CED-approved studies, and 116 (49 percent) were non-CED-approved studies. Compared with non-CED-approved studies, CED-approved studies had more robust designs (67 percent versus 13 percent randomized clinical trials), larger enrollment (median sample size of 1,000 versus 125), and more frequent inclusion of US patients (93 percent versus 44 percent). CMS's reliance on non-CED-approved studies for CED reconsiderations, despite CED-approved studies being higher quality and more representative of Medicare beneficiaries, highlights opportunities to strengthen the CED program.