Background: Major obstacles in global control of TB are inadequate ability to rapidly and accurately diagnose active TB in developing countries. All over the world, data of notified cases to national TB control program show that most of TB diagnosis could not be confirmed bacteriologically, and thus proportion of these clinically diagnosed TB cases might not have had TB. Sheep Blood Agar (SBA) is used in routine microbiology laboratories for the purpose of diagnosis in different bacterial infectious diseases, and so it is important to evaluate for the culture of M.tuberculosis. Aim and objective: The objective of this study was to evaluate the performance of 7% SBA for the primary isolation of Mycobacterium tuberculosis from pulmonary specimens to determine the possibility of using 7% SBA for diagnosis of TB in resource limited settings and so as to improve the diagnostic potential of 7% SBA by adding cheap antimicrobial agents. Methodology: Health-facility based comparative crossectional study was conducted from 212 tuberculosis suspected individuals in Addis Ababa, Ethiopia during November, 2013 to March, 2014. Sputum samples were analyzed using AFB smear microscopy, LJ and 7% SBA culture. PCR- based molecular characterization of RD9 deletion typing and Mycobacterium genus typing was done for AFB confirmed isolates and compared the cost of performing per specimen culture analyses using these two media. Result: The Sensitivity, specificity, positive predictive value and negative predictive value of blood agar media compared with golden standard of LJ was 96.4%, 98%, 94.7% and 98.7%, respectively. There was significant difference in the detection days of macroscopic colonies in the media with a median days to detect macroscopic colonies on between 7% SBA and LJ were 14 days (Interquartile range 13-19) and 20 days (Interquartile range 14-23) respectively (P<0.0001). The rate of contamination was on SBA 20/424 (4.7 %) and on LJ 22/424 (5.2%). The current result of this study show that, although the actual medium was cheap for 7% SBA and LJ methods, the cost per culture (specimen) in consumables and labour cost is comparable between the two methods in USD 3.76 (USD 3760/1000 specimens) for 7% SBA and USD 4.01 (USD $ 4010/1000 specimens) for LJ. Conclusion: SBA medium may be a good alternative to LJ medium for rapid detection of M.TB from sputum in resource-limited settings. ZN staining of sediment using 4% NaOH may be more alternative for urgent management of PTB patients than culture methods. SBA cost is comparable to LJ and may save about one fourth of the detection time in contrast to LJ.