Patients with diabetes had twice the prevalence of latent tuberculosis infection (LTBI) compared to those without diabetes in cross-sectional studies in the United States. However, it was unknown whether LTBI increased the risk of diabetes. Therefore, researchers used longitudinal data to see if LTBI was linked to an increased risk of diabetes. From 2000 to 2015, investigators conducted a retrospective cohort study among U.S. Veterans receiving care from the Veterans Health Administration. All patients without diabetes who had a tuberculin skin test (TST) or an interferon-γ release assay (IGRA) were eligible. Patients with active tuberculosis and those diagnosed with diabetes before or within 2 years of LTBI testing were excluded. Patients were tracked until they were diagnosed with diabetes, died, or reached the year 2015. TST or IGRA positivity was used to define LTBI. ICD-9 numbers, in conjunction with a diabetic medicine prescription, were used to characterize incident diabetes. Of the 574,113 individuals that were eligible, 5.3% had both TST and IGRA, 79.1% had TST solely, and 15.6% had IGRA only. Overall, 6.6% of people tested positive for LTBI, with 2,535,149 person-years (PY) of follow-up after the test (median 3.2 years). Patients with LTBI had a higher diabetes incidence rate (per 100,000 PY) than those without (1,012 vs 744; hazard ratio [HR] 1.4 [95% CI 1.3–1.4]). After controlling for confounders, the increased diabetes incidence remained (aHR 1.2 [95% CI 1.2–1.3]) as compared to those who did not have LTBI. Diabetes incidence was similar in patients with LTBI who were treated for it versus those who were not (aHR 1.0 [95% CI 0.9–1.1]). Longitudinal research suggested that LTBI was linked to an increased risk of diabetes. The outcomes have ramifications for the 25% of the world’s population who suffer from LTBI.