In a population with a high prevalence of tuberculosis, no statistically significant risk of developing pulmonary tuberculosis could be attributed to the long-term use of systemic corticosteroid therapy in a group of asthmatic men. A trend towards an increased risk of tuberculosis was noted but the extent of the attributable risk was too small to justify routine treatment with isoniazid. There is certainly no justification for withholding necessary corticosteroid therapy in black subjects with asthma.