Introduction. - We evaluated the utility of the TB-info® software in the management of patients receiving antituberculosis treatment (ATT). Methods. - A cohort study of patients who started TAT in 2004 and were followed up at two hospitals of the Assistance Publique des Hôpitaux de Paris and at five medical-social centers of the City of Paris using the TB-Info® software. These results were compared with those previously published for the 1996–2003 cohort. Results. - Two hundred and nine TAT regimens were initiated in 2004, with an average duration of 7.2 months. The epidemiological and clinical characteristics reflect the precarious situation of the study population. Compared to the previous cohort, screening for viral co-infections was performed in over 80% of cases: 11%, 10%, and 5% of patients were co-infected with HIV, HBV, or HCV, respectively. In this cohort, only 21 patients were lost to follow-up (11.2%), and the success rate for directly observed positive pulmonary forms was 76%, with 34% of cases showing evidence of cure confirmed by end-of-treatment samples. Conclusion. - Analysis of data from the TB-info® software, in accordance with WHO recommendations on the documentation of treatment outcomes, shows an improvement in clinical practices: increased screening for co-infections and a decrease in the number of patients lost to follow-up. A gap remains with international targets. These results confirm the value of the IT tool for patient follow-up and the evaluation of professional practices in France.