The major aim is to develop and test various educational strategies targeting TB-infected adolescents at two health centers in Los Angeles in an experimental design to assess the relative effectiveness on medication adherence, appointment keeping, and completion of therapy. The educational content and modalities are based on an intervention-oriented model of behavioral action derived from the theory of reasoned action, the theory of planned behavior, and the transtheoretical model. An experimental four group design will be used to assess the independent and combined effects of peer counseling, a participant/parent contingency contract intervention, and a combination of peer counseling and contingency contracting, compared with a usual care control group. A total of 880 participants will be randomly assigned to one of the four groups. In addition to assessing the impact of the intervention, analyses will compare the relative importance of the hypothesized influencing and mediating factors on behavior. Cost-effectiveness analyses will also be conducted. During the later stages of the study, clinic staff from throughout Los Angeles County will be trained and encouraged to integrate the strategies that appear to be effective into their clinic setting.