We conducted a population-based serosurvey of urban areas and rural regions of southern Ghana, West Africa. Subjects (3763) of all ages were enrolled from 25 city and village sites and in studies of groups of special interest. "Positive" results were difficult to define because of a high frequency of results that were indeterminate on immunoblotting, the current standard for confirmation of HTLV-I. However, polymerase chain reaction results and HTLV type-specific discriminatory tests proved HTLV-I was present in Ghana. No HTLV-2 positivity was observed. By using strict criteria that considered indeterminate results as negative, the overall prevalence was found to be between 1 and 2% in all areas, with no difference by geographic location. Prevalence rose with age and was higher in adult women than men. However, in substudies of selected populations, we found HTLV prevalence among 124 persons with lymphomas and hematological malignancies was not different from that in the general population. Furthermore, the prevalence in prostitutes was similar to that in the general population and in pregnant women. HTLV-I is present in West Africa, but we were unable to associate HTLV-I seropositivity with malignancy or with prostitution.