Living donor kidney transplants (LDKTs) are the primary reason for racial/ethnic disparities in kidney transplantation in recent years – between 2000 and 2010, whites obtained LDKTs at rates more than double that of African Americans. Why is this the case? One possibility, which I call the “differential access” hypothesis, argues that this is an instance of health disparities begetting health disparities – since kin are the most common types of donors and are frequently of the same race, African Americans’ potential donor pool is sicker than whites’, leading to lower levels of access to medically suitable living donors. However, research testing this hypothesis has only examined the donor evaluation outcomes of potential donors, not the characteristics of members of the transplant candidate’s social network that could be evaluated, but are not. Closing this data gap is critical to understanding the role of candidates’ networks in disparities in LDKT. Furthermore, in the last decade, the number of LDKTs has steadily decreased even as the waiting list has grown past 100,000 candidates and the number of deceased donor kidney transplants has increased at a rate insufficient to keep up with demand. In the medium term, LDKTs are arguably the most promising avenue by which to combat this trend – but only if more patients have access to medically suitable and potentially willing living donors than are currently obtaining LDKTs. Therefore, measuring transplant candidates’ social network characteristics to assess prospective donor suitability is a critical step to measuring the degree of underutilization of LDKT for all racial groups, and for designing effective and ethical interventions to promote higher rates of LDKT. I report on the results of a pilot study of 72 transplant candidates and their relationships with 1,611 members of their social networks. These data do not support the differential access hypothesis – if anything, African American transplant candidates appear to have access to more suitable donors than whites. Instead, white candidates appear to be advantaged by a higher rate of prospective donors approaching them to be evaluated without being asked. As candidates rarely decline such offers in these data, promoting potential donor volunteerism and eliminating barriers to doing so seems to be a promising point of intervention. We also find that 86% of candidates have access to at least one promising prospective donor in their networks, suggesting that the extent of underutilization is large. This appears to be especially pronounced among more distant relatives such as nieces/nephews, cousins, and aunts/uncles.