BACKGROUND The presence of a genetic factor in the determination of leprosy has long been debated. This study tests whether the HLA-linked control of susceptibility to leprosy and/or for the types of leprosy could be confirmed. METHODS In 15 multicase families, the method of DeVries et al., 1976, was used to detect nonrandom segregation of parental HLA haplotypes in their affected and healthy siblings. Linkage analyses, for two and three alleles were performed by the computer program LIPED: RESULTS For the affected siblings, the segregations of the parental HLA haplotype were significantly nonrandom from the healthy parents and random from the affected parents, indicating that affected siblings were sharing their HLA haplotypes (segregated from the healthy parents) more than expected. The segregations to the healthy siblings from both the healthy and affected parents were random. Healthy siblings inherited the haplotypes shared among the leprosy siblings randomly as expected. There were excess DR2/DR2 homozygote individuals among tuberculoid siblings. The highest lod score was achieved when we considered our suggested three-alleles model for the susceptibility to the different types of leprosy. CONCLUSIONS A closely HLA-linked gene on chromosome number 6 with multiple alleles (3 or more) in recombination fraction between 0.05 and 0.1 with 70 to 100% penetrance may be responsible for the susceptibility to the different types of leprosy, whereas the susceptibility to leprosy per se maybe the responsibility of non-HLA linked gene/s. DR2/DR2 homozygote individuals may be relatively at high risk of developing leprosy or tuberculoid leprosy.