Background: Jails are considered reservoirs of Tb, therefore, an important threat to public health. Objective: to determine the incidence and epidemiology of Tb in prisoners in four jails, two of each for male and female inmates. Methods: Prospective cohort study. All prisoners with respiratory symptoms and their personal contacts were evaluated. After signed consent forms, we collected three spontaneous sputum and blood samples on consecutive days. Auramine-rodhamine stain and cultures in thin layer agar, Löwestein-Jensen and MGIT were performed. Additionally, susceptibility tests were performed for first line drugs and HIV testing. We also collected socio-demographic and clinical data, and personal and family history. To determine overcrowding conditions, the width and length of the cells were measured and number of prisoners in each cell was determined. Results: We evaluated 2103 prisoners, and identified 961 with lower respiratory symptoms (92.5% were men). Among them, 47 were diagnosed with Tb, 12 (25.5%) were detected only by culture. The annual incidence was 494 cases/100000 prisoners (Medellin: 515/100000; Bucaramanga: 443/100000). The median age was 30 years, 25.5% (12) had less than 15 days of cough, 12.7% had past history of TB and 44.6% (21) had prior contact with a Tb patient, 13 of the contacts were prisoners. TB-HIV co-infection was diagnosed in two cases, and we found two cases resistant to streptomycin and one to isoniazid. The median of square meters per person was 1.84 (IQR = 0.96-3.56). Conclusion: TB in jails was very high in our study, about 20 times higher than in general population in Colombia that is 25 cases/100000 inhabitants. This illness should be considered in inmate patients with lower respiratory symptoms of any duration confined in those places and perform mycobacterial culture.