Background: Tuberculosis is one of the most ancient recognised diseases of human being which remains as an important obstacle for public health, despite numerous scientific advances made in the recent decades. Delay in treatment of tuberculosis due to socieoeconomic factors in addition to HIV epidemic increases the number of undiagnosed smear positive patients in communities. The patient’s delay and physician’s delay in diagnosis and treatment, inadequate therapy, errors in prescribing accurate regimens, financial problems of patients in taking drugs, lack of patient’s knowledge about the regular use of drugs and interruption of treatment all result in the emergence of drug resistant tuberculosis which adds to the difficulties in the management of this obstacle. Regarding the difficulties faced in treatment of tuberculosis, determining factors responsible for treatment failure and the consequently developing resistance is essential. Materials and Methods: This is a case control and sequential study which consists of hospitalised and out patient TB cases who have remained smear positive despite full course of anti-TB treatment. Eighty patients with this characteristic were compared with 80 controls who have responded to treatment. Results: Eighty cases (44 females, 36 males) with median age of 28 yr. (15-73) were compared with eighty controls (41 females, 39 males) with median age of 29 yr. (16-70). There were no differences between two groups regarding nationality (Iranian vs. Afghan) (p>0.05) and economic status (p>0.05). There was significance difference regarding educational level of two groups, as it was significantly higher in control group (p <0.05). There was no significant statistical difference between two groups when living place was considered (urban, suburb, village) (p>0.05). The final interesting findings were high usage rate of four drug regimen and DOTS strategy in control group in comparison with the patients with treatment failure. (p <0.05, p <0.05, respectively). If the time from the onset of patients symptoms to the initiation of proper treatment (total delay) was more than 8 weeks, no significant difference was noted. However, if this time period was less than 8 weeks, significant difference was detected (p <0.05). This study shows that an important factor in preventing from treatment failure and emergence of resistant forms is the immediate referral of the patient to physician (less than 8 weeks) resulting in prompt diagnosis and treatment. Conclusion: Finally, DOTS strategy, early diagnosis, correct treatment of tuberculosis, patient’s knowledge about tuberculosis, how to use and side effects of anti-TB drugs and duration of treatment all should be considered as effective factors for prevention of anti tuberculosis treatment failure and emergence of drug resistant bacilli. (Tanaffos 2003; 2 (7): 61-68)