Mycobacterium tuberculosis infection remains a major health problem throughout the world. In France, tuberculosis is endemic, particularly in the Paris area (Ile-de-France) and in the south (Provence Alpes côte d'Azur) where immigration is greater than in other countries in northern Europe. Culture is the gold standard for diagnosis of tuberculosis and is the only method enabling a study of strain sensitivity to treatment. Histology contributes to diagnosis in most cases by revealing typical necrotizing granulomatous lesions. The diagnosis is then confirmed by the detection of acid-fast bacilli with Ziehl-Neelsen staining. However, the Ziehl-Neelsen stain is not sensitive and does not allow identification of different species. The polymerase chain reaction (PCR) DNA amplification method has been used to detect M. tuberculosis in formalin-fixed paraffin-embedded tissues. The aim of the present study was to investigate the value of this method for the diagnosis of M. tuberculosis infection. The results obtained with PCR assay were compared to those obtained with histological and microbiological methods (direct examination and culture). Sixty-three specimens (mainly lymph node and lung specimens) exhibiting a positive culture for M. tuberculosis were analyzed. Tuberculosis granulomas were noted in 32/63 cases, tuberculoid granulomas in 18/63, pyoepitheloid granuloms in 10/63, and non-specific inflammation in 3/63. Ziehl-Neelsen staining was positive in 11/63 cases. PCR assay on tissue sections was positive for M. tuberculosis in 58/63 cases. Controls of the PCR method (granulomas due to other mycobacterial species, foreign body granulomas, sarcoidosis granulomas) were all negative. This study shows that PCR from deparaffinized sections, 1) greatly increases the sensitivity of diagnosis of tuberculosis, 2) enables the diagnosis of M. tuberculosis infection. However, although this method reduces the time to diagnosis, culture remains the gold standard for identification of the mycobacterium and for determining the sensitivity of the isolated strain to treatment.