Background: Study done in Ethiopia in Southern Nation Nationality People (SNNP) during 2003 showed the incidence of TB among those on HAART was lower (3.70 cases of TB per 100 person years) than those, on pre-HAART (11.1 cases of TB per 100 PYO). Besides being on HAART or pre HAART different socio demographic, substance use and clinical factors play an important role in developing TB infection among PLHIV. So, the objective of this study was to assess the effect of HAART on incidence of TB among people living with HIV/AIDS. 1.2 Methods: A retrospective cohort study was conducted at Assela referral Hospital among patients enrolled on HIV care and support during September 13/2005 to January 30/2011. A total of 412 patients' record from the HAART and pre HAART group were selected in one to one ratio by simple random sampling method. Kaplan Meier and proportional cox regression methods were used to determine the predictor of TB incidence.Result: The incidence of tuberculosis was 7.02/100 person years (95% CI: 5.02, 9.83) among those on pre HAART follow up where as it was 3.73/100 person years among those on HAART. Furthermore, individuals on HAART have a 96.8% decrease in risk of tuberculosis, (AHR=0.032; 95% CI: 0.012, 0.082) than those non-HAART individuals. Similarly CD4 cell count <200 cell/µl and WHO clinical stage III or IV has statistically significant association with TB development among People living with HIV/AIDS. Conclusion and recommendation:HAART use decreased tuberculosis incidence among HIV positive individuals. In addition, CD4 cell count <200 cell/µl and WHO clinical stage III or IV were factors associated with the development of new TB cases among PLHIV. So, HAART should be started with a higher CD4 cell count and none advanced WHO clinical stage in order to get maximum reduction of new TB cases among PLHIV.