Background: The tuberculosis control strategy of vaccinating First Nations newborns with BCG (bacille Calmette-Guerin) is currently undergoing re-evaluation in Canada. Review of recent pediatric tuberculosis morbidity could inform this re-evaluation. Methods: Potential source cases and pediatric cases of tuberculosis from Alberta First Nations were identified over the 10 years 1991–2000. The distribution of pediatric disease was described. The effect of BCG on tuberculosis morbidity in two large outbreaks was determined. Results: A total of 57 potential source cases and 41 pediatric cases of tuberculosis were reported from 17 (41.5%) and 8 (19.5%) of the 41 on-reserve First Nation Community Health Centres, respectively. Three outbreaks traceable to three source cases accounted for 34 (18, 3, and 13, respectively) of the 41 (82.9%) pediatric cases. Each outbreak was spatially and temporally separate from the other. Each outbreak strain of Mycobacterium tuberculosis had a unique DNA fingerprint. In the largest outbreaks, disease-to-infection ratios (secondary case rates) were higher in newly infected unvaccinated versus vaccinated close pediatric contacts (12/13 [92.3%] versus 7/15 [46.7%], p=0.02), but the infection rate was almost certainly falsely high in the BCG vaccinated. One unvaccinated child had a brain tuberculoma in addition to primary pulmonary tuberculosis. Conclusion: For most Alberta First Nations communities, the spatial and temporal distribution of disease, and the meager impact on morbidity, challenge the rationale for continued use of BCG.