A recent study suggested that a dodecamer duplication in exon 42 of the HOPA gene in Xq13 may be a significant factor in the etiology of X-linked mental retardation. In an effort to investigate this possibility, we determined the incidence of the dodecamer duplication in cohorts of non-fragile X males with mental retardation from three countries, cohorts of fragile X males from two countries, 43 probands from families with X-linked mental retardation and control cohorts from three countries. The duplication was found in 3.6-4.0% of male patients from two non-fragile X groups (Italy and South Carolina), in 1.2% from another non-fragile X group (South Africa), but in no male patients from families with X-linked mental retardation (South Carolina). The dodecamer duplication was also found in several white males with fragile X syndrome from France (5%) and South Africa (22.2%). Additionally, the duplication was found in 1.5% of South Carolinian newborn males, 2.5% South Carolinian male college students, 5% Italian male controls and 4.5% of the white South African controls. None of the black South African non-fragile X individuals with mental retardation, the fragile X or the control samples tested carried the duplication, suggesting that the duplication is rare in the black South African population. The incidence of the duplication was not significantly different between any of the groups in the study. Therefore, results of our studies in four different populations do not corroborate the findings of the previous study, and indicate that the HOPA dodecamer duplication does not convey an increased susceptibility to mental retardation.