A health institution based cross-sectional studyrnwas carried out in Shashemene town, Southern Ethiopiarnbetween September, 1993, and January, 1994, to determinernthe sero-prevalence and the clinical impact of HIV amongrnnewly diagnosed tuberculosis patients.rnThe HIV-antibody was determined using the two ELISArnprocedures (Wellcozyme, Wellcome Diagnostics, DartfordrnKent England, and Du-pont assay, Singapore).rnA total of 450 tuberculosis patients aged 15 yearsrnand above were enrolled in the study. The overall HIVseroprevalencernrate was 44.4%. The highest rate wasrnobserved in the age group 20-39 years.rnA slightly higher HIV-infection rate was found inrnmales (46%) than in females (41%). Those divorced andrnwidowed patients had higher proportion of HIV seropositivity.rnThe HIV positivity rate was higher forrnextra-pulmonary than pulmonary form of tuberculosis (ORrn= 3.80; 95% CI: 1.49, 9.7 ). Higher proportion of sputumrnpositive patients were HIV-positive compared to thernsputum negative pulmonary tuberculosis patients (OR=l. 09;rn95% CI: 0.64,1.85) though they manifested typicalrnradiographic features (OR = 7.87; 95% CI: 4. 39, 14.21) .rnvirnSignificant differences were noted among HIVrnpositives than HIV negatives in manifesting herpesrnzoster, lymphadenopathy, oral candidiasis, peripheralrnparaesthesia and chronic diarrhoea .rnAn alternat i ve diagnostic approach is required tornavoid missing HIV-related tuberculosis which isrncommunicable to the general population. Other findingsrnare discussed and recommendations made.