OBJECTIVE: To determine the association, if any, between the presence of oral lesions and clinical and immunological status of untreated HIV-infected adults in Tanzania. DESIGN: A cross-sectional study. SETTING: AIDS Clinical Trial Clinic (ATCC) at Muhimbili Medical Centre in Dar-es-Salaam, Tanzania. SUBJECTS: 192 HIV-infected individuals not receiving treatment; 156 individuals confirmed to be HIV-seronegative acted as a control group. METHODS: Examination of oral structures, determination of HIV serostatus, clinical status, and peripheral CD4 ؉ T cell and total lymphocyte counts. MAIN OUTCOME MEASURE: Presence of oral lesions. RESULTS: Intra-oral lesions were seen among 7.7% of the HIV-seronegative, 10.4% of the HIV-seropositive and 36.8% of the AIDS groups, respectively. Enlarged parotid glands were seen in 20% of the AIDS patients, 11.9% of the HIV-seropositives, and 5.1% of the HIV seronegatives. Enlargement of submandibular salivary glands was seen in 29.6% of the AIDS patients, 31.3% of the HIV-seropositives compared with 14.7% among the HIV-seronegatives. Multiple regression analysis was used to calculate adjusted odds ratio (OR) for presence of oral lesions. OR for an intra-oral lesion was 1.6 (95% CI ؍ 0.5; 5.0) among the HIV-seropositives and 8.2 (95% CI ؍ 3.5; 19.7) among the AIDS patients using the HIV-seronegatives as reference. OR for an intra-oral lesion was 0.9 (95% CI ؍ 0.3; 2.9) in HIV-infected patients with peripheral CD4 ؉ T cell count of between 200-500 cells mm ؊3 and 2.7 (95% CI ؍ 0.9; 7.7) in patients with less than 200 cells mm ؊3 . OR for an intra-oral lesion was 0.4 (95% CI ؍ 0.2; 0.9) for patients with peripheral total lymphocyte counts of cells mm ؊3 and 0.9 (95 CI ؍ 0.4; 2.0) for patients with less than 1000 cells mm ؊3 . CONCLUSION: The association of oral lesions with the clinical stage of HIV infection and to a lesser extent peripheral CD4 ؉ T cell count does suggest that these lesions could be used as additional markers of immunosuppression and AIDS.