This study evaluated the diagnostic accuracy of the WHO decision-making algorithms for the treatment of tuberculosis in children with suspected pulmonary TB. 1,886 children under 10 years of age were evaluated, of whom 20.3% had HIV and 15.1% were severely malnourished. The TDA A algorithm (with chest X-ray) had a sensitivity of 84.3% and a specificity of 50.6%. The TDA B algorithm (without X-ray) had a sensitivity of 90.6% and a specificity of 30.8%. In children in the high-risk groups, the sensitivity was lower than in the low-risk children, but the specificity was higher. The results confirmed that the algorithms have high sensitivity but low specificity. Further prospective studies and integration of new diagnostic tools are needed to improve accuracy, especially specificity.