Introduction: The risk of tuberculosis is greatly increased in those with HIV infection, but the relative risk will vary over time, depending on the proportion of HIV-infected individuals with different levels of immunosuppression. This study in Chiang Rai. Thailand. assesses how this relative risk changes, investigates any interaction with age. and estimates the TB burden attributable to HIV infection now and in the future. Methods: We conducted a case-control study using retrospective data from Chiang Rai hospital. Cases were all newly diagnosed TB patients during 1990-1998. Controls were antenatal clinic attenders (ANC), delivery patients, surgical patients, blood donors and military conscripts. Odds ratios (OR) were calculated separately by year, age group and gender. using each control group separately. The population attributable fraction (PAF) was calculated by year. A mathematical model was developed to recalculate the PAF taking account of the time lag between HIV infection and TB, and to predict PAF in the future. Results: During the study period, the number of new TB cases in Chiang Rai hospital increased more than 3-fold. HIV prevalence peaked at 17.4% in 1992 in male military recruits before declining dramatically. The peak was later and lower in women. The OR for the association of TB and HIV infection increased markedly over time but there was no consistent pattern by age. Age-adjusted PAFs, rose to around 70% by 1998. Modelling suggested,that a true difference in ORs by age is masked by the lower proportion of individuals with more advanced HIV infection in younger age groups. The model gave slightly lower estimates for the PAF than those calculated directly, and predicted that the PAF would decline earlier in younger individuals. For older adults PAFs will remain very high throughout this decade. Conclusion: The HIV epidemic has a profound and prolonged impact on TB burden despite the marked reduction in HIV incidence already seen in Chiang Rai. Besides HIV prevention, we need additional TB control measures to reduce the burden among HIV positives.