BACKGROUND: Basic science, epidemiological and interventional research supports a link between vitamin D and tuberculosis (TB) immunity, infection and disease. We evaluated the association between vitamin D levels and TB infection and disease in UK children recruited to the National Institute for Health Research IGRA Kids Study (NIKS).METHODS: Children presenting between 2011 and 2014 were eligible if they had history of exposure to an adult case with sputum smear/culture-positive TB, or were referred and diagnosed with TB disease. Children were assessed at baseline and at 6-8 weeks for immunological evidence of TB infection (interferon-gamma release assay and/or tuberculin skin test) and evidence of TB disease. Some centres routinely measured total 25-hydroxy vitamin D (25-OHD) levels.RESULTS: A total of 166 children were included. The median 25-OHD levels were higher in non-infected children (45.5 nmol/l) than in those with tuberculous infection (36.2 nmol/l) and TB disease (20.0 nmol/l). The difference between TB infection and disease was statistically significant (P < 0.001). By logistic regression, lower vitamin D levels were associated with TB disease among participants with infection or disease, with no evidence of confounding by age, sex, bacille Calmette-Guérin vaccination status, ethnicity, non-contact referral, season or centre.CONCLUSION: Children with TB disease had lower vitamin D levels than children with infection. Implications for prevention and treatment remain to be established.