Indochinese refugees entering the United States have a high rate of tuberculosis and tuberculin reactivity. In addition, several investigators have noted that a large number of refugees with initial tuberculin tests that are "not significant" change to "significant" reactions when retested within 8 wk. This "conversion" phenomenon has been reported in 21 to 43% of refugees and has been unexplained by antigen, testing, demographic, or exposure risk factors. A prospective evaluation of 218 refugees, conducted to assess the role of anergy and boosting, confirmed earlier findings, with 52% of 118 persons with initial tuberculin reactions that were "not significant" developing "significant" reactions on subsequent testing. Anergy, as measured by nonreactivity to mumps and candida skin tests, was not found to be a contributing factor, as few refugees were anergic and as rates of anergy did not differ significantly among refugees with different responses to tuberculin. Boosting, however, played a major role in explaining the "conversions," as 59% of persons who changed to "significant" tuberculin tests did so when retested with tuberculin at 1 to 3 wk. "Delayed" boosting rather than incubating disease or anergy appeared to be the most likely explanation for the remaining "conversions" that occurred on a third PPD test conducted at approximately 8 wk. If the "conversion" phenomenon is due to boosting, it remains to be seen whether the boosting is a result of previous exposure to Mycobacterium tuberculosis or to other, nontuberculous mycobacteria.