In 1989, a population‐based cohort of persons aged ≥50 years was established in an urban area of Guinea‐Bissau, West Africa. Overall, 346 persons were interviewed in detail about risk behaviors and had capillary blood drawn. Among women, 12.4% were HTLV‐1 seropositive, compared with 4.6% in men. No HTLV‐2 was found. Seropositivity varied considerably according to place of birth and ethnic group. In women, but not in men, HTLV‐1 seropositivity was strongly associated with early sexual debut (10–14 yrs, 33.3%; 15–17 yrs, 26.0%; 18–20 yrs, 6.5%; 21+ yrs, 0%; ptrend = 0.001), lifetime number of male partners (ptrend = 0.006), and the male partner's number of co‐wives (ptrend = 0.006). There was also a 3.1‐fold increased risk of being HTLV‐1 seropositive if the woman was also HIV‐2 seropositive. In a multivariate‐risk‐factor analysis, the strongest association with HTLV‐1 was a history of having been bitten by a monkey (n = 11; combined ORadjusted = 10.1; 95% CI 2.3–44.4). Ornamental scarification was associated with a 3.3‐fold increased risk. Ethnic affiliation also significantly influenced the risk of being HTLV‐1 seropositive. Follow‐up performed in January 1996 revealed no difference in survival between HTLV‐1‐seropositive and ‐seronegative individuals over 6 years (rate ratio = 1.4, 95% CI 0.7–2.8). In conclusion, this population, which has very high HIV‐2 seroprevalence, is also highly endemic for HTLV‐1. Whereas sexual behaviors are clearly important for HTLV‐1 spread in women, non‐sexual risk factors were the only ones of potential importance in men. HTLV‐1 had no impact on survival in this older population. Int. J. Cancer 76:293–298, 1998.© 1998 Wiley‐Liss, Inc.