There is convincing evidence that host genes affect the outcome of infection in human tuberculosis. Two complementary strategies were used to identify the genes involved. A linkage-based genome-wide screen was carried out to locate the positions of genes exerting a major population-wide effect on tuberculosis susceptibility. A candidate-gene-based case-control study was used to examine the effects of genes that may exert a more moderate effect on risk of clinical tuberculosis. The genome screen was conducted in two stages. In the first stage 299 microsatellite markers, spanning all 23 chromosomes, were typed in 92 independent sib-pairs, and seven regions showed some evidence of co-segregation with the disease. These seven regions were examined in a second set of 81 sib-pairs, and markers on chromosomes 15q and Xq showed evidence of linkage to tuberculosis. An X chromosome susceptibility gene may contribute to the excess of males with tuberculosis observed in many populations. The candidate gene approach compared the frequency of polymorphisms in several genes in over 400 subjects with smear-positive pulmonary tuberculosis and 400 ethnically matched healthy controls. Polymorphisms in genes encoding natural-resistance-associated macrophage protein, vitamin D receptor and mannose-binding lectin were associated with tuberculosis. These results suggest that many genes may be involved in determining host susceptibility to tuberculosis, and highlight the importance of using several different study methods to locate them.