Accurately identifying latent tuberculosis (TB) infection (LTBI) in liver and renal transplant candidates is important because of the risks associated with both treatment of LTBI and reactivation of disease in this population. Many programs advocate yearly screening of patients awaiting organ transplantation. The reproducibility of serial interferon-gamma release assay (IGRA) testing in transplant candidates has not been studied. We conducted a retrospective longitudinal study of patients listed for liver or kidney transplantation between January 1, 2005 and February 1, 2012 at the University of Illinois Medical Center at Chicago. Data collected included demographics, transplant type, IGRA results, treatment received, and mortality. The study population was comprised of 795 adults; 79 (10%) had at least 1 indeterminate result; indeterminate results were less common in men (P = 0.01) and more common in liver transplant candidates (P < 0.001). The reversion frequency was 27% with a rate of 158.1 reversions in 1000 person-years. A higher magnitude of initial TB response values was predictive of consistently positive results (P < 0.001). The conversion frequency was 15% with a rate of 82.6 conversions in 1000 person-years. Among those who converted, the values of the IGRA varied, with 48% having a TB response of <1 IU/mL, 41% 1-5 IU/mL, and only 10% >5 IU/mL. A significant number of conversions and reversions occur during serial IGRA testing of transplant candidates. Delineating true-positive converters from false-positives is an issue that warrants further study. © 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd.