Only a few series of pediatric tuberculosis (TB) have been reported in the last 20 years. The purpose of this study was to evaluate the clinical, radiological, microbiological, and treatment characteristics of childhood TB. A total of 539 children with childhood TB diagnosed over a 12-year period (1994-2005) in 16 different centers in Turkey participated in the study. The medical records of all childhood TB patients were investigated. A total of 539 children (274 males, 265 females) with childhood TB aged 10 days-17 years participated in the study. Age distribution was nearly equal among all age groups. We detected the index case in 39.8% of the patients. More than one index case was detected in 17.3% of the patients. A minimum 15-mm induration is accepted on tuberculin skin test (TST) following Bacillus Calmette-Guérin (BCG) vaccination. The TST was positive in 55.3% of the patients. Acid-fast bacillus smear was positive in 133, and polymerase chain reaction for Mycobacterium tuberculosis was positive in 45 patients. In 75 patients (13.9%), cultures yielded M. tuberculosis. One hundred fifty-one patients (28%) did not present for followup, and families of 5 patients (0.9%) discontinued the treatment. Pulmonary TB (n=285) and meningeal TB (n=85) were the most frequent diseases. In 29% of the patients, there was poor adherence to treatment or patients were lost to follow-up. We have demonstrated that household contact screening procedures play a major and important role, especially considering the high ratio of cases with contact index cases. We also recommend that the positive TST values should be reviewed according to the local cut-off data and should be specified in as many countries as possible. In view of the considerably high percentages of patients lost to follow-up and treatment discontinuation observed in our study, we suggest that application of directly observed treatment short-course (DOTS) is preferable.