TB is typically caused by Mycobacterium tuberculosis, a symbiotic bacterium present in one-third of the world's population. There any many factors triggering overt clinical disease in a small proportion of humans. In our view the major role in the process is played by the host's immune response, especially self-directed, destructive inflammation. Conventional chemotherapy produces bactericidal or bacteriostatic effects, but immunopathological changes can only be corrected by immunotherapy. Various attempts have been made to identify the optimal immune intervention. Some have shown promising effects, but many have failed. It is commonly believed that the field started in 1890: the year Robert Koch announced his tuberculin therapy. In the Pên Ts'ao Kang Mu, classical Chinese materia medica, published during Ming dynasty, Li Shi Chen (1518–1593) recommended, as a remedy for hemoptysis, to collect from the sputum “…blood lumps, roast them till they are black, and take then them as a powder”. In retrospect, this is perhaps the earliest recorded reference relating to immunotherapy of TB with heat-killed mycobacteria. Modern science is obviously geared toward more palatable approach, but without hindsight from often disdained empirical evidence no progress can be made. The clinical experience from various trial and error processes is briefly discussed in this review.